{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.08921004505107274,
  "eval_steps": 1000,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 225.1666717529297,
      "epoch": 8.921004505107276e-05,
      "grad_norm": 62.695068359375,
      "learning_rate": 2.5e-07,
      "loss": 12.7202,
      "reward": 0.19306249171495438,
      "reward_std": 0.5882241576910019,
      "rewards/correctness_reward_func": 0.3333333432674408,
      "rewards/int_reward_func": 0.0833333358168602,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.22360415756702423,
      "step": 1,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.00017842009010214551,
      "grad_norm": 59.6503791809082,
      "learning_rate": 5e-07,
      "loss": 13.0732,
      "step": 2
    },
    {
      "epoch": 0.00026763013515321824,
      "grad_norm": 69.65516662597656,
      "learning_rate": 7.5e-07,
      "loss": 12.9681,
      "step": 3
    },
    {
      "epoch": 0.00035684018020429103,
      "grad_norm": 57.81648635864258,
      "learning_rate": 1e-06,
      "loss": 8.1042,
      "step": 4
    },
    {
      "epoch": 0.00044605022525536376,
      "grad_norm": 57.6408576965332,
      "learning_rate": 1.25e-06,
      "loss": 8.6056,
      "step": 5
    },
    {
      "epoch": 0.0005352602703064365,
      "grad_norm": 58.459903717041016,
      "learning_rate": 1.5e-06,
      "loss": 10.4929,
      "step": 6
    },
    {
      "epoch": 0.0006244703153575092,
      "grad_norm": 62.41658020019531,
      "learning_rate": 1.7500000000000002e-06,
      "loss": 13.1206,
      "step": 7
    },
    {
      "epoch": 0.0007136803604085821,
      "grad_norm": 66.22370910644531,
      "learning_rate": 2e-06,
      "loss": 13.2007,
      "step": 8
    },
    {
      "epoch": 0.0008028904054596548,
      "grad_norm": 66.21946716308594,
      "learning_rate": 2.25e-06,
      "loss": 12.3522,
      "step": 9
    },
    {
      "epoch": 0.0008921004505107275,
      "grad_norm": 65.43058776855469,
      "learning_rate": 2.5e-06,
      "loss": 7.9566,
      "step": 10
    },
    {
      "epoch": 0.0009813104955618004,
      "grad_norm": 54.532962799072266,
      "learning_rate": 2.75e-06,
      "loss": 8.8616,
      "step": 11
    },
    {
      "epoch": 0.001070520540612873,
      "grad_norm": 56.53645706176758,
      "learning_rate": 3e-06,
      "loss": 10.3095,
      "step": 12
    },
    {
      "completion_length": 222.9791717529297,
      "epoch": 0.0011597305856639458,
      "grad_norm": 78.4708023071289,
      "learning_rate": 3e-06,
      "loss": -24.6031,
      "reward": 0.011666670441627502,
      "reward_std": 0.524684801697731,
      "rewards/correctness_reward_func": 0.1250000037252903,
      "rewards/int_reward_func": 0.041666666977107525,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.1550000049173832,
      "step": 13,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.0012489406307150184,
      "grad_norm": 88.30101776123047,
      "learning_rate": 3e-06,
      "loss": -18.3145,
      "step": 14
    },
    {
      "epoch": 0.0013381506757660913,
      "grad_norm": 101.97128295898438,
      "learning_rate": 3e-06,
      "loss": -7.4151,
      "step": 15
    },
    {
      "epoch": 0.0014273607208171641,
      "grad_norm": 91.58382415771484,
      "learning_rate": 3e-06,
      "loss": -8.9073,
      "step": 16
    },
    {
      "epoch": 0.0015165707658682367,
      "grad_norm": 90.10670471191406,
      "learning_rate": 3e-06,
      "loss": -13.5176,
      "step": 17
    },
    {
      "epoch": 0.0016057808109193096,
      "grad_norm": 80.67254638671875,
      "learning_rate": 3e-06,
      "loss": -17.2813,
      "step": 18
    },
    {
      "epoch": 0.0016949908559703822,
      "grad_norm": 75.51331329345703,
      "learning_rate": 3e-06,
      "loss": -24.6926,
      "step": 19
    },
    {
      "epoch": 0.001784200901021455,
      "grad_norm": 78.15167999267578,
      "learning_rate": 3e-06,
      "loss": -18.5973,
      "step": 20
    },
    {
      "epoch": 0.0018734109460725277,
      "grad_norm": 89.70745086669922,
      "learning_rate": 3e-06,
      "loss": -7.9364,
      "step": 21
    },
    {
      "epoch": 0.0019626209911236007,
      "grad_norm": 89.28164672851562,
      "learning_rate": 3e-06,
      "loss": -9.434,
      "step": 22
    },
    {
      "epoch": 0.002051831036174673,
      "grad_norm": 98.30489349365234,
      "learning_rate": 3e-06,
      "loss": -14.6494,
      "step": 23
    },
    {
      "epoch": 0.002141041081225746,
      "grad_norm": 92.3221206665039,
      "learning_rate": 3e-06,
      "loss": -17.5654,
      "step": 24
    },
    {
      "completion_length": 224.95834350585938,
      "epoch": 0.002230251126276819,
      "grad_norm": 77.60931396484375,
      "learning_rate": 3e-06,
      "loss": 12.7299,
      "reward": 0.08664583414793015,
      "reward_std": 0.529650554060936,
      "rewards/correctness_reward_func": 0.1666666679084301,
      "rewards/int_reward_func": 0.041666666977107525,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.12168749794363976,
      "step": 25,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.0023194611713278916,
      "grad_norm": 55.48664474487305,
      "learning_rate": 3e-06,
      "loss": 11.3297,
      "step": 26
    },
    {
      "epoch": 0.0024086712163789645,
      "grad_norm": 64.88197326660156,
      "learning_rate": 3e-06,
      "loss": 7.6398,
      "step": 27
    },
    {
      "epoch": 0.002497881261430037,
      "grad_norm": 66.41521453857422,
      "learning_rate": 3e-06,
      "loss": 10.9742,
      "step": 28
    },
    {
      "epoch": 0.0025870913064811097,
      "grad_norm": 60.356266021728516,
      "learning_rate": 3e-06,
      "loss": 18.3629,
      "step": 29
    },
    {
      "epoch": 0.0026763013515321826,
      "grad_norm": 67.53816986083984,
      "learning_rate": 3e-06,
      "loss": 10.3122,
      "step": 30
    },
    {
      "epoch": 0.0027655113965832554,
      "grad_norm": 81.81299591064453,
      "learning_rate": 3e-06,
      "loss": 12.4031,
      "step": 31
    },
    {
      "epoch": 0.0028547214416343282,
      "grad_norm": 58.01384735107422,
      "learning_rate": 3e-06,
      "loss": 11.3115,
      "step": 32
    },
    {
      "epoch": 0.0029439314866854006,
      "grad_norm": 60.38798522949219,
      "learning_rate": 3e-06,
      "loss": 7.5438,
      "step": 33
    },
    {
      "epoch": 0.0030331415317364735,
      "grad_norm": 76.68485260009766,
      "learning_rate": 3e-06,
      "loss": 9.8314,
      "step": 34
    },
    {
      "epoch": 0.0031223515767875463,
      "grad_norm": 63.667381286621094,
      "learning_rate": 3e-06,
      "loss": 18.0907,
      "step": 35
    },
    {
      "epoch": 0.003211561621838619,
      "grad_norm": 64.93324279785156,
      "learning_rate": 3e-06,
      "loss": 9.6529,
      "step": 36
    },
    {
      "completion_length": 200.9166717529297,
      "epoch": 0.003300771666889692,
      "grad_norm": 55.603302001953125,
      "learning_rate": 3e-06,
      "loss": -2.1381,
      "reward": 0.20900000631809235,
      "reward_std": 0.5408279597759247,
      "rewards/correctness_reward_func": 0.2500000074505806,
      "rewards/int_reward_func": 0.09374999813735485,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.1347499955445528,
      "step": 37,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.0033899817119407644,
      "grad_norm": 59.070777893066406,
      "learning_rate": 3e-06,
      "loss": -6.7825,
      "step": 38
    },
    {
      "epoch": 0.0034791917569918372,
      "grad_norm": 73.52457427978516,
      "learning_rate": 3e-06,
      "loss": -11.5592,
      "step": 39
    },
    {
      "epoch": 0.00356840180204291,
      "grad_norm": 68.8139419555664,
      "learning_rate": 3e-06,
      "loss": -3.9847,
      "step": 40
    },
    {
      "epoch": 0.003657611847093983,
      "grad_norm": 74.64259338378906,
      "learning_rate": 3e-06,
      "loss": -7.7023,
      "step": 41
    },
    {
      "epoch": 0.0037468218921450553,
      "grad_norm": 68.76261901855469,
      "learning_rate": 3e-06,
      "loss": -11.4536,
      "step": 42
    },
    {
      "epoch": 0.003836031937196128,
      "grad_norm": 57.10056686401367,
      "learning_rate": 3e-06,
      "loss": -3.0195,
      "step": 43
    },
    {
      "epoch": 0.003925241982247201,
      "grad_norm": 57.4798583984375,
      "learning_rate": 3e-06,
      "loss": -7.3677,
      "step": 44
    },
    {
      "epoch": 0.004014452027298274,
      "grad_norm": 62.251949310302734,
      "learning_rate": 3e-06,
      "loss": -12.4481,
      "step": 45
    },
    {
      "epoch": 0.004103662072349346,
      "grad_norm": 67.0556640625,
      "learning_rate": 3e-06,
      "loss": -4.2431,
      "step": 46
    },
    {
      "epoch": 0.0041928721174004195,
      "grad_norm": 79.22687530517578,
      "learning_rate": 3e-06,
      "loss": -8.9896,
      "step": 47
    },
    {
      "epoch": 0.004282082162451492,
      "grad_norm": 83.6895980834961,
      "learning_rate": 3e-06,
      "loss": -12.4645,
      "step": 48
    },
    {
      "completion_length": 186.0,
      "epoch": 0.004371292207502565,
      "grad_norm": 187.8282928466797,
      "learning_rate": 3e-06,
      "loss": -44.8522,
      "reward": 0.5453333556652069,
      "reward_std": 0.9364342093467712,
      "rewards/correctness_reward_func": 0.4583333358168602,
      "rewards/int_reward_func": 0.15625,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.06924999551847577,
      "step": 49,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.004460502252553638,
      "grad_norm": 92.9270248413086,
      "learning_rate": 3e-06,
      "loss": -41.2773,
      "step": 50
    },
    {
      "epoch": 0.00454971229760471,
      "grad_norm": 212.18917846679688,
      "learning_rate": 3e-06,
      "loss": -42.3882,
      "step": 51
    },
    {
      "epoch": 0.004638922342655783,
      "grad_norm": 102.22235870361328,
      "learning_rate": 3e-06,
      "loss": -42.8879,
      "step": 52
    },
    {
      "epoch": 0.004728132387706856,
      "grad_norm": 79.1269302368164,
      "learning_rate": 3e-06,
      "loss": -44.708,
      "step": 53
    },
    {
      "epoch": 0.004817342432757929,
      "grad_norm": 94.53079986572266,
      "learning_rate": 3e-06,
      "loss": -41.656,
      "step": 54
    },
    {
      "epoch": 0.004906552477809001,
      "grad_norm": 91.7303695678711,
      "learning_rate": 3e-06,
      "loss": -45.3257,
      "step": 55
    },
    {
      "epoch": 0.004995762522860074,
      "grad_norm": 92.66773986816406,
      "learning_rate": 3e-06,
      "loss": -41.4113,
      "step": 56
    },
    {
      "epoch": 0.005084972567911147,
      "grad_norm": 123.76467895507812,
      "learning_rate": 3e-06,
      "loss": -43.4643,
      "step": 57
    },
    {
      "epoch": 0.0051741826129622194,
      "grad_norm": 109.21142578125,
      "learning_rate": 3e-06,
      "loss": -44.7136,
      "step": 58
    },
    {
      "epoch": 0.005263392658013293,
      "grad_norm": 83.24272155761719,
      "learning_rate": 3e-06,
      "loss": -45.7862,
      "step": 59
    },
    {
      "epoch": 0.005352602703064365,
      "grad_norm": 94.45966339111328,
      "learning_rate": 3e-06,
      "loss": -42.5492,
      "step": 60
    },
    {
      "completion_length": 225.625,
      "epoch": 0.0054418127481154375,
      "grad_norm": 125.85611724853516,
      "learning_rate": 3e-06,
      "loss": 28.3825,
      "reward": 0.31822918355464935,
      "reward_std": 0.9613562524318695,
      "rewards/correctness_reward_func": 0.375,
      "rewards/int_reward_func": 0.1145833320915699,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.17135417833924294,
      "step": 61,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.005531022793166511,
      "grad_norm": 128.64669799804688,
      "learning_rate": 3e-06,
      "loss": 4.8596,
      "step": 62
    },
    {
      "epoch": 0.005620232838217583,
      "grad_norm": 287.9391784667969,
      "learning_rate": 3e-06,
      "loss": 20.5521,
      "step": 63
    },
    {
      "epoch": 0.0057094428832686565,
      "grad_norm": 111.01509857177734,
      "learning_rate": 3e-06,
      "loss": 16.6241,
      "step": 64
    },
    {
      "epoch": 0.005798652928319729,
      "grad_norm": 123.25679016113281,
      "learning_rate": 3e-06,
      "loss": 6.8919,
      "step": 65
    },
    {
      "epoch": 0.005887862973370801,
      "grad_norm": 115.68987274169922,
      "learning_rate": 3e-06,
      "loss": 19.3061,
      "step": 66
    },
    {
      "epoch": 0.0059770730184218746,
      "grad_norm": 128.9923553466797,
      "learning_rate": 3e-06,
      "loss": 27.4792,
      "step": 67
    },
    {
      "epoch": 0.006066283063472947,
      "grad_norm": 130.64230346679688,
      "learning_rate": 3e-06,
      "loss": 3.8702,
      "step": 68
    },
    {
      "epoch": 0.00615549310852402,
      "grad_norm": 169.2925262451172,
      "learning_rate": 3e-06,
      "loss": 19.2163,
      "step": 69
    },
    {
      "epoch": 0.006244703153575093,
      "grad_norm": 104.88905334472656,
      "learning_rate": 3e-06,
      "loss": 14.5854,
      "step": 70
    },
    {
      "epoch": 0.006333913198626165,
      "grad_norm": 134.32022094726562,
      "learning_rate": 3e-06,
      "loss": 5.6117,
      "step": 71
    },
    {
      "epoch": 0.006423123243677238,
      "grad_norm": 124.52132415771484,
      "learning_rate": 3e-06,
      "loss": 18.0908,
      "step": 72
    },
    {
      "completion_length": 203.81250762939453,
      "epoch": 0.006512333288728311,
      "grad_norm": 87.01981353759766,
      "learning_rate": 3e-06,
      "loss": 14.5961,
      "reward": 0.25443750619888306,
      "reward_std": 0.6893003582954407,
      "rewards/correctness_reward_func": 0.2916666641831398,
      "rewards/int_reward_func": 0.09374999813735485,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.13097916916012764,
      "step": 73,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.006601543333779384,
      "grad_norm": 83.70246887207031,
      "learning_rate": 3e-06,
      "loss": 20.7203,
      "step": 74
    },
    {
      "epoch": 0.006690753378830456,
      "grad_norm": 80.23466491699219,
      "learning_rate": 3e-06,
      "loss": 30.3319,
      "step": 75
    },
    {
      "epoch": 0.006779963423881529,
      "grad_norm": 74.68209838867188,
      "learning_rate": 3e-06,
      "loss": 23.2,
      "step": 76
    },
    {
      "epoch": 0.006869173468932602,
      "grad_norm": 81.28849029541016,
      "learning_rate": 3e-06,
      "loss": 11.7216,
      "step": 77
    },
    {
      "epoch": 0.0069583835139836745,
      "grad_norm": 85.60411071777344,
      "learning_rate": 3e-06,
      "loss": 19.9348,
      "step": 78
    },
    {
      "epoch": 0.007047593559034747,
      "grad_norm": 95.26403045654297,
      "learning_rate": 3e-06,
      "loss": 13.5735,
      "step": 79
    },
    {
      "epoch": 0.00713680360408582,
      "grad_norm": 81.69352722167969,
      "learning_rate": 3e-06,
      "loss": 19.4906,
      "step": 80
    },
    {
      "epoch": 0.0072260136491368926,
      "grad_norm": 80.9581527709961,
      "learning_rate": 3e-06,
      "loss": 29.1989,
      "step": 81
    },
    {
      "epoch": 0.007315223694187966,
      "grad_norm": 87.37995147705078,
      "learning_rate": 3e-06,
      "loss": 23.4541,
      "step": 82
    },
    {
      "epoch": 0.007404433739239038,
      "grad_norm": 90.7470932006836,
      "learning_rate": 3e-06,
      "loss": 10.7907,
      "step": 83
    },
    {
      "epoch": 0.007493643784290111,
      "grad_norm": 352.26953125,
      "learning_rate": 3e-06,
      "loss": 18.1423,
      "step": 84
    },
    {
      "completion_length": 185.6875,
      "epoch": 0.007582853829341184,
      "grad_norm": 78.5768051147461,
      "learning_rate": 3e-06,
      "loss": -2.0743,
      "reward": 0.6054166778922081,
      "reward_std": 0.8349271714687347,
      "rewards/correctness_reward_func": 0.5,
      "rewards/int_reward_func": 0.125,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.019583335146307945,
      "step": 85,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.007672063874392256,
      "grad_norm": 125.65332794189453,
      "learning_rate": 3e-06,
      "loss": -18.0183,
      "step": 86
    },
    {
      "epoch": 0.00776127391944333,
      "grad_norm": 93.01673889160156,
      "learning_rate": 3e-06,
      "loss": -2.9219,
      "step": 87
    },
    {
      "epoch": 0.007850483964494403,
      "grad_norm": 85.38358306884766,
      "learning_rate": 3e-06,
      "loss": -7.3606,
      "step": 88
    },
    {
      "epoch": 0.007939694009545474,
      "grad_norm": 99.59243774414062,
      "learning_rate": 3e-06,
      "loss": -18.9376,
      "step": 89
    },
    {
      "epoch": 0.008028904054596548,
      "grad_norm": 96.83404541015625,
      "learning_rate": 3e-06,
      "loss": -7.9748,
      "step": 90
    },
    {
      "epoch": 0.008118114099647621,
      "grad_norm": 81.16954803466797,
      "learning_rate": 3e-06,
      "loss": -4.2134,
      "step": 91
    },
    {
      "epoch": 0.008207324144698692,
      "grad_norm": 123.15869140625,
      "learning_rate": 3e-06,
      "loss": -19.8823,
      "step": 92
    },
    {
      "epoch": 0.008296534189749766,
      "grad_norm": 93.05419158935547,
      "learning_rate": 3e-06,
      "loss": -4.5813,
      "step": 93
    },
    {
      "epoch": 0.008385744234800839,
      "grad_norm": 106.2331314086914,
      "learning_rate": 3e-06,
      "loss": -8.6969,
      "step": 94
    },
    {
      "epoch": 0.00847495427985191,
      "grad_norm": 99.65939331054688,
      "learning_rate": 3e-06,
      "loss": -21.3275,
      "step": 95
    },
    {
      "epoch": 0.008564164324902984,
      "grad_norm": 94.40375518798828,
      "learning_rate": 3e-06,
      "loss": -9.7937,
      "step": 96
    },
    {
      "completion_length": 217.25000762939453,
      "epoch": 0.008653374369954057,
      "grad_norm": 133.5598907470703,
      "learning_rate": 3e-06,
      "loss": -68.7329,
      "reward": 0.6968958526849747,
      "reward_std": 0.7409922480583191,
      "rewards/correctness_reward_func": 0.5833333283662796,
      "rewards/int_reward_func": 0.1979166641831398,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.08435416780412197,
      "step": 97,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.00874258441500513,
      "grad_norm": 136.60848999023438,
      "learning_rate": 3e-06,
      "loss": -74.5256,
      "step": 98
    },
    {
      "epoch": 0.008831794460056202,
      "grad_norm": 123.70120239257812,
      "learning_rate": 3e-06,
      "loss": -59.6774,
      "step": 99
    },
    {
      "epoch": 0.008921004505107275,
      "grad_norm": 150.22532653808594,
      "learning_rate": 3e-06,
      "loss": -69.5624,
      "step": 100
    },
    {
      "epoch": 0.009010214550158348,
      "grad_norm": 126.68507385253906,
      "learning_rate": 3e-06,
      "loss": -62.8973,
      "step": 101
    },
    {
      "epoch": 0.00909942459520942,
      "grad_norm": 105.47962951660156,
      "learning_rate": 3e-06,
      "loss": -61.6182,
      "step": 102
    },
    {
      "epoch": 0.009188634640260493,
      "grad_norm": 144.26048278808594,
      "learning_rate": 3e-06,
      "loss": -70.5109,
      "step": 103
    },
    {
      "epoch": 0.009277844685311567,
      "grad_norm": 141.22325134277344,
      "learning_rate": 3e-06,
      "loss": -76.5479,
      "step": 104
    },
    {
      "epoch": 0.009367054730362638,
      "grad_norm": 139.37173461914062,
      "learning_rate": 3e-06,
      "loss": -62.353,
      "step": 105
    },
    {
      "epoch": 0.009456264775413711,
      "grad_norm": 150.77801513671875,
      "learning_rate": 3e-06,
      "loss": -72.2384,
      "step": 106
    },
    {
      "epoch": 0.009545474820464785,
      "grad_norm": 138.2374267578125,
      "learning_rate": 3e-06,
      "loss": -65.3746,
      "step": 107
    },
    {
      "epoch": 0.009634684865515858,
      "grad_norm": 132.50453186035156,
      "learning_rate": 3e-06,
      "loss": -64.141,
      "step": 108
    },
    {
      "completion_length": 198.125,
      "epoch": 0.00972389491056693,
      "grad_norm": 187.5413055419922,
      "learning_rate": 3e-06,
      "loss": 44.1489,
      "reward": 0.7788957953453064,
      "reward_std": 0.7549726963043213,
      "rewards/correctness_reward_func": 0.6666666567325592,
      "rewards/int_reward_func": 0.1770833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.06485416181385517,
      "step": 109,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.009813104955618003,
      "grad_norm": 138.89434814453125,
      "learning_rate": 3e-06,
      "loss": 51.6646,
      "step": 110
    },
    {
      "epoch": 0.009902315000669076,
      "grad_norm": 128.95484924316406,
      "learning_rate": 3e-06,
      "loss": 31.7993,
      "step": 111
    },
    {
      "epoch": 0.009991525045720148,
      "grad_norm": 126.7931900024414,
      "learning_rate": 3e-06,
      "loss": 38.5454,
      "step": 112
    },
    {
      "epoch": 0.01008073509077122,
      "grad_norm": 125.33599853515625,
      "learning_rate": 3e-06,
      "loss": 40.3822,
      "step": 113
    },
    {
      "epoch": 0.010169945135822294,
      "grad_norm": 139.41482543945312,
      "learning_rate": 3e-06,
      "loss": 32.5052,
      "step": 114
    },
    {
      "epoch": 0.010259155180873366,
      "grad_norm": 169.09432983398438,
      "learning_rate": 3e-06,
      "loss": 43.5542,
      "step": 115
    },
    {
      "epoch": 0.010348365225924439,
      "grad_norm": 133.872802734375,
      "learning_rate": 3e-06,
      "loss": 50.3469,
      "step": 116
    },
    {
      "epoch": 0.010437575270975512,
      "grad_norm": 125.77018737792969,
      "learning_rate": 3e-06,
      "loss": 31.112,
      "step": 117
    },
    {
      "epoch": 0.010526785316026585,
      "grad_norm": 128.32257080078125,
      "learning_rate": 3e-06,
      "loss": 36.629,
      "step": 118
    },
    {
      "epoch": 0.010615995361077657,
      "grad_norm": 124.38401794433594,
      "learning_rate": 3e-06,
      "loss": 39.8284,
      "step": 119
    },
    {
      "epoch": 0.01070520540612873,
      "grad_norm": 138.24668884277344,
      "learning_rate": 3e-06,
      "loss": 31.1433,
      "step": 120
    },
    {
      "completion_length": 191.6041717529297,
      "epoch": 0.010794415451179804,
      "grad_norm": 199.1646270751953,
      "learning_rate": 3e-06,
      "loss": 86.314,
      "reward": 1.0014583468437195,
      "reward_std": 0.8148851096630096,
      "rewards/correctness_reward_func": 0.875,
      "rewards/int_reward_func": 0.2291666641831398,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.10270833596587181,
      "step": 121,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.010883625496230875,
      "grad_norm": 195.7254638671875,
      "learning_rate": 3e-06,
      "loss": 76.1289,
      "step": 122
    },
    {
      "epoch": 0.010972835541281948,
      "grad_norm": 175.13900756835938,
      "learning_rate": 3e-06,
      "loss": 86.1448,
      "step": 123
    },
    {
      "epoch": 0.011062045586333022,
      "grad_norm": 182.21661376953125,
      "learning_rate": 3e-06,
      "loss": 90.0805,
      "step": 124
    },
    {
      "epoch": 0.011151255631384093,
      "grad_norm": 189.17214965820312,
      "learning_rate": 3e-06,
      "loss": 76.0951,
      "step": 125
    },
    {
      "epoch": 0.011240465676435166,
      "grad_norm": 195.55718994140625,
      "learning_rate": 3e-06,
      "loss": 89.5242,
      "step": 126
    },
    {
      "epoch": 0.01132967572148624,
      "grad_norm": 171.1396484375,
      "learning_rate": 3e-06,
      "loss": 82.3705,
      "step": 127
    },
    {
      "epoch": 0.011418885766537313,
      "grad_norm": 189.04995727539062,
      "learning_rate": 3e-06,
      "loss": 71.8677,
      "step": 128
    },
    {
      "epoch": 0.011508095811588384,
      "grad_norm": 162.9297332763672,
      "learning_rate": 3e-06,
      "loss": 81.2432,
      "step": 129
    },
    {
      "epoch": 0.011597305856639458,
      "grad_norm": 173.23104858398438,
      "learning_rate": 3e-06,
      "loss": 85.8069,
      "step": 130
    },
    {
      "epoch": 0.011686515901690531,
      "grad_norm": 162.6637420654297,
      "learning_rate": 3e-06,
      "loss": 69.8347,
      "step": 131
    },
    {
      "epoch": 0.011775725946741603,
      "grad_norm": 190.06675720214844,
      "learning_rate": 3e-06,
      "loss": 84.5222,
      "step": 132
    },
    {
      "completion_length": 207.33334350585938,
      "epoch": 0.011864935991792676,
      "grad_norm": 147.16957092285156,
      "learning_rate": 3e-06,
      "loss": -63.4697,
      "reward": 0.6791666746139526,
      "reward_std": 1.0425111949443817,
      "rewards/correctness_reward_func": 0.5833333134651184,
      "rewards/int_reward_func": 0.1770833358168602,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.08124999818392098,
      "step": 133,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.011954146036843749,
      "grad_norm": 154.6719970703125,
      "learning_rate": 3e-06,
      "loss": -70.5112,
      "step": 134
    },
    {
      "epoch": 0.01204335608189482,
      "grad_norm": 137.32408142089844,
      "learning_rate": 3e-06,
      "loss": -41.1322,
      "step": 135
    },
    {
      "epoch": 0.012132566126945894,
      "grad_norm": 126.37704467773438,
      "learning_rate": 3e-06,
      "loss": -53.2367,
      "step": 136
    },
    {
      "epoch": 0.012221776171996967,
      "grad_norm": 152.24891662597656,
      "learning_rate": 3e-06,
      "loss": -58.4567,
      "step": 137
    },
    {
      "epoch": 0.01231098621704804,
      "grad_norm": 116.28028106689453,
      "learning_rate": 3e-06,
      "loss": -46.2973,
      "step": 138
    },
    {
      "epoch": 0.012400196262099112,
      "grad_norm": 152.08795166015625,
      "learning_rate": 3e-06,
      "loss": -62.8325,
      "step": 139
    },
    {
      "epoch": 0.012489406307150185,
      "grad_norm": 146.10671997070312,
      "learning_rate": 3e-06,
      "loss": -71.6559,
      "step": 140
    },
    {
      "epoch": 0.012578616352201259,
      "grad_norm": 149.14556884765625,
      "learning_rate": 3e-06,
      "loss": -42.1534,
      "step": 141
    },
    {
      "epoch": 0.01266782639725233,
      "grad_norm": 151.06182861328125,
      "learning_rate": 3e-06,
      "loss": -55.6968,
      "step": 142
    },
    {
      "epoch": 0.012757036442303403,
      "grad_norm": 145.29530334472656,
      "learning_rate": 3e-06,
      "loss": -60.2759,
      "step": 143
    },
    {
      "epoch": 0.012846246487354477,
      "grad_norm": 124.00696563720703,
      "learning_rate": 3e-06,
      "loss": -48.5856,
      "step": 144
    },
    {
      "completion_length": 200.14583587646484,
      "epoch": 0.012935456532405548,
      "grad_norm": 104.97675323486328,
      "learning_rate": 3e-06,
      "loss": -22.2294,
      "reward": 0.21922918409109116,
      "reward_std": 0.6296879947185516,
      "rewards/correctness_reward_func": 0.2916666716337204,
      "rewards/int_reward_func": 0.062499999068677425,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.13493750616908073,
      "step": 145,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.013024666577456621,
      "grad_norm": 83.18937683105469,
      "learning_rate": 3e-06,
      "loss": -22.3765,
      "step": 146
    },
    {
      "epoch": 0.013113876622507695,
      "grad_norm": 96.22801971435547,
      "learning_rate": 3e-06,
      "loss": -22.5564,
      "step": 147
    },
    {
      "epoch": 0.013203086667558768,
      "grad_norm": 102.87374877929688,
      "learning_rate": 3e-06,
      "loss": -24.9001,
      "step": 148
    },
    {
      "epoch": 0.01329229671260984,
      "grad_norm": 110.96674346923828,
      "learning_rate": 3e-06,
      "loss": -18.8972,
      "step": 149
    },
    {
      "epoch": 0.013381506757660913,
      "grad_norm": 91.87604522705078,
      "learning_rate": 3e-06,
      "loss": -18.1615,
      "step": 150
    },
    {
      "epoch": 0.013470716802711986,
      "grad_norm": 88.4422836303711,
      "learning_rate": 3e-06,
      "loss": -23.0431,
      "step": 151
    },
    {
      "epoch": 0.013559926847763058,
      "grad_norm": 83.86327362060547,
      "learning_rate": 3e-06,
      "loss": -23.25,
      "step": 152
    },
    {
      "epoch": 0.01364913689281413,
      "grad_norm": 82.81922149658203,
      "learning_rate": 3e-06,
      "loss": -23.1331,
      "step": 153
    },
    {
      "epoch": 0.013738346937865204,
      "grad_norm": 104.8452377319336,
      "learning_rate": 3e-06,
      "loss": -26.8428,
      "step": 154
    },
    {
      "epoch": 0.013827556982916276,
      "grad_norm": 92.94257354736328,
      "learning_rate": 3e-06,
      "loss": -20.0667,
      "step": 155
    },
    {
      "epoch": 0.013916767027967349,
      "grad_norm": 84.95638275146484,
      "learning_rate": 3e-06,
      "loss": -19.1472,
      "step": 156
    },
    {
      "completion_length": 231.7916717529297,
      "epoch": 0.014005977073018422,
      "grad_norm": 139.96688842773438,
      "learning_rate": 3e-06,
      "loss": 21.1797,
      "reward": 0.6720625460147858,
      "reward_std": 0.9181468784809113,
      "rewards/correctness_reward_func": 0.5833333432674408,
      "rewards/int_reward_func": 0.2187500074505806,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.13002083078026772,
      "step": 157,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.014095187118069494,
      "grad_norm": 109.121337890625,
      "learning_rate": 3e-06,
      "loss": 10.5428,
      "step": 158
    },
    {
      "epoch": 0.014184397163120567,
      "grad_norm": 94.9039306640625,
      "learning_rate": 3e-06,
      "loss": 4.8126,
      "step": 159
    },
    {
      "epoch": 0.01427360720817164,
      "grad_norm": 109.7251968383789,
      "learning_rate": 3e-06,
      "loss": 5.2961,
      "step": 160
    },
    {
      "epoch": 0.014362817253222714,
      "grad_norm": 103.42703247070312,
      "learning_rate": 3e-06,
      "loss": 4.0648,
      "step": 161
    },
    {
      "epoch": 0.014452027298273785,
      "grad_norm": 127.93770599365234,
      "learning_rate": 3e-06,
      "loss": 7.0101,
      "step": 162
    },
    {
      "epoch": 0.014541237343324858,
      "grad_norm": 145.8150634765625,
      "learning_rate": 3e-06,
      "loss": 18.3559,
      "step": 163
    },
    {
      "epoch": 0.014630447388375932,
      "grad_norm": 116.2653579711914,
      "learning_rate": 3e-06,
      "loss": 8.3424,
      "step": 164
    },
    {
      "epoch": 0.014719657433427003,
      "grad_norm": 104.55130767822266,
      "learning_rate": 3e-06,
      "loss": 2.084,
      "step": 165
    },
    {
      "epoch": 0.014808867478478076,
      "grad_norm": 114.84294128417969,
      "learning_rate": 3e-06,
      "loss": 2.2571,
      "step": 166
    },
    {
      "epoch": 0.01489807752352915,
      "grad_norm": 99.8189468383789,
      "learning_rate": 3e-06,
      "loss": 0.9219,
      "step": 167
    },
    {
      "epoch": 0.014987287568580221,
      "grad_norm": 142.80715942382812,
      "learning_rate": 3e-06,
      "loss": 3.505,
      "step": 168
    },
    {
      "completion_length": 214.77083587646484,
      "epoch": 0.015076497613631295,
      "grad_norm": 114.3720703125,
      "learning_rate": 3e-06,
      "loss": -32.1329,
      "reward": 1.0511458218097687,
      "reward_std": 1.0028848350048065,
      "rewards/correctness_reward_func": 0.9166666865348816,
      "rewards/int_reward_func": 0.3125,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.17802083492279053,
      "step": 169,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.015165707658682368,
      "grad_norm": 100.3597412109375,
      "learning_rate": 3e-06,
      "loss": -38.7511,
      "step": 170
    },
    {
      "epoch": 0.015254917703733441,
      "grad_norm": 108.30574035644531,
      "learning_rate": 3e-06,
      "loss": -46.3083,
      "step": 171
    },
    {
      "epoch": 0.015344127748784513,
      "grad_norm": 116.34545135498047,
      "learning_rate": 3e-06,
      "loss": -39.7363,
      "step": 172
    },
    {
      "epoch": 0.015433337793835586,
      "grad_norm": 113.52851104736328,
      "learning_rate": 3e-06,
      "loss": -34.5686,
      "step": 173
    },
    {
      "epoch": 0.01552254783888666,
      "grad_norm": 110.65509796142578,
      "learning_rate": 3e-06,
      "loss": -32.8796,
      "step": 174
    },
    {
      "epoch": 0.01561175788393773,
      "grad_norm": 107.06590270996094,
      "learning_rate": 3e-06,
      "loss": -32.6552,
      "step": 175
    },
    {
      "epoch": 0.015700967928988806,
      "grad_norm": 100.2861557006836,
      "learning_rate": 3e-06,
      "loss": -39.6106,
      "step": 176
    },
    {
      "epoch": 0.015790177974039876,
      "grad_norm": 107.69467163085938,
      "learning_rate": 3e-06,
      "loss": -46.9244,
      "step": 177
    },
    {
      "epoch": 0.01587938801909095,
      "grad_norm": 96.8420181274414,
      "learning_rate": 3e-06,
      "loss": -40.93,
      "step": 178
    },
    {
      "epoch": 0.015968598064142022,
      "grad_norm": 113.12389373779297,
      "learning_rate": 3e-06,
      "loss": -37.0258,
      "step": 179
    },
    {
      "epoch": 0.016057808109193095,
      "grad_norm": 116.10971069335938,
      "learning_rate": 3e-06,
      "loss": -34.9046,
      "step": 180
    },
    {
      "completion_length": 218.64583587646484,
      "epoch": 0.01614701815424417,
      "grad_norm": 307.6673889160156,
      "learning_rate": 3e-06,
      "loss": -29.0457,
      "reward": 1.4360832571983337,
      "reward_std": 1.0610616505146027,
      "rewards/correctness_reward_func": 1.25,
      "rewards/int_reward_func": 0.34375,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.15766665898263454,
      "step": 181,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.016236228199295242,
      "grad_norm": 125.99212646484375,
      "learning_rate": 3e-06,
      "loss": 1.016,
      "step": 182
    },
    {
      "epoch": 0.01632543824434631,
      "grad_norm": 112.21588897705078,
      "learning_rate": 3e-06,
      "loss": -17.115,
      "step": 183
    },
    {
      "epoch": 0.016414648289397385,
      "grad_norm": 118.06622314453125,
      "learning_rate": 3e-06,
      "loss": -0.4864,
      "step": 184
    },
    {
      "epoch": 0.016503858334448458,
      "grad_norm": 116.36631774902344,
      "learning_rate": 3e-06,
      "loss": 3.4437,
      "step": 185
    },
    {
      "epoch": 0.01659306837949953,
      "grad_norm": 124.60052490234375,
      "learning_rate": 3e-06,
      "loss": -27.5515,
      "step": 186
    },
    {
      "epoch": 0.016682278424550605,
      "grad_norm": 160.65628051757812,
      "learning_rate": 3e-06,
      "loss": -29.3863,
      "step": 187
    },
    {
      "epoch": 0.016771488469601678,
      "grad_norm": 127.9763412475586,
      "learning_rate": 3e-06,
      "loss": 0.7423,
      "step": 188
    },
    {
      "epoch": 0.01686069851465275,
      "grad_norm": 116.69316101074219,
      "learning_rate": 3e-06,
      "loss": -18.6518,
      "step": 189
    },
    {
      "epoch": 0.01694990855970382,
      "grad_norm": 114.2183609008789,
      "learning_rate": 3e-06,
      "loss": -0.87,
      "step": 190
    },
    {
      "epoch": 0.017039118604754894,
      "grad_norm": 126.1614761352539,
      "learning_rate": 3e-06,
      "loss": 2.8213,
      "step": 191
    },
    {
      "epoch": 0.017128328649805968,
      "grad_norm": 134.43527221679688,
      "learning_rate": 3e-06,
      "loss": -28.8518,
      "step": 192
    },
    {
      "completion_length": 247.08334350585938,
      "epoch": 0.01721753869485704,
      "grad_norm": 111.3412094116211,
      "learning_rate": 3e-06,
      "loss": -7.8814,
      "reward": 1.0157291293144226,
      "reward_std": 0.7945153564214706,
      "rewards/correctness_reward_func": 0.9166666865348816,
      "rewards/int_reward_func": 0.2916666716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.1926041767001152,
      "step": 193,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.017306748739908114,
      "grad_norm": 116.42599487304688,
      "learning_rate": 3e-06,
      "loss": -15.1245,
      "step": 194
    },
    {
      "epoch": 0.017395958784959187,
      "grad_norm": 136.37391662597656,
      "learning_rate": 3e-06,
      "loss": -9.2426,
      "step": 195
    },
    {
      "epoch": 0.01748516883001026,
      "grad_norm": 97.36872863769531,
      "learning_rate": 3e-06,
      "loss": -10.8671,
      "step": 196
    },
    {
      "epoch": 0.01757437887506133,
      "grad_norm": 125.0397720336914,
      "learning_rate": 3e-06,
      "loss": -7.0755,
      "step": 197
    },
    {
      "epoch": 0.017663588920112404,
      "grad_norm": 171.17971801757812,
      "learning_rate": 3e-06,
      "loss": -16.797,
      "step": 198
    },
    {
      "epoch": 0.017752798965163477,
      "grad_norm": 100.81266021728516,
      "learning_rate": 3e-06,
      "loss": -8.4577,
      "step": 199
    },
    {
      "epoch": 0.01784200901021455,
      "grad_norm": 127.79389953613281,
      "learning_rate": 3e-06,
      "loss": -16.3874,
      "step": 200
    },
    {
      "epoch": 0.017931219055265624,
      "grad_norm": 131.9748077392578,
      "learning_rate": 3e-06,
      "loss": -10.7973,
      "step": 201
    },
    {
      "epoch": 0.018020429100316697,
      "grad_norm": 100.95606231689453,
      "learning_rate": 3e-06,
      "loss": -12.0026,
      "step": 202
    },
    {
      "epoch": 0.018109639145367767,
      "grad_norm": 131.19261169433594,
      "learning_rate": 3e-06,
      "loss": -8.3155,
      "step": 203
    },
    {
      "epoch": 0.01819884919041884,
      "grad_norm": 164.74656677246094,
      "learning_rate": 3e-06,
      "loss": -18.9275,
      "step": 204
    },
    {
      "completion_length": 191.7291717529297,
      "epoch": 0.018288059235469913,
      "grad_norm": 150.95191955566406,
      "learning_rate": 3e-06,
      "loss": 50.6719,
      "reward": 1.3118958473205566,
      "reward_std": 0.8902758955955505,
      "rewards/correctness_reward_func": 0.9999999701976776,
      "rewards/int_reward_func": 0.34375,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.03185417060740292,
      "step": 205,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.018377269280520987,
      "grad_norm": 128.34344482421875,
      "learning_rate": 3e-06,
      "loss": 32.3302,
      "step": 206
    },
    {
      "epoch": 0.01846647932557206,
      "grad_norm": 136.15789794921875,
      "learning_rate": 3e-06,
      "loss": 33.8857,
      "step": 207
    },
    {
      "epoch": 0.018555689370623133,
      "grad_norm": 140.50901794433594,
      "learning_rate": 3e-06,
      "loss": 28.924,
      "step": 208
    },
    {
      "epoch": 0.018644899415674206,
      "grad_norm": 168.0647430419922,
      "learning_rate": 3e-06,
      "loss": 31.3019,
      "step": 209
    },
    {
      "epoch": 0.018734109460725276,
      "grad_norm": 133.79208374023438,
      "learning_rate": 3e-06,
      "loss": 31.6401,
      "step": 210
    },
    {
      "epoch": 0.01882331950577635,
      "grad_norm": 161.34898376464844,
      "learning_rate": 3e-06,
      "loss": 49.2047,
      "step": 211
    },
    {
      "epoch": 0.018912529550827423,
      "grad_norm": 129.22007751464844,
      "learning_rate": 3e-06,
      "loss": 30.2,
      "step": 212
    },
    {
      "epoch": 0.019001739595878496,
      "grad_norm": 143.37449645996094,
      "learning_rate": 3e-06,
      "loss": 31.2762,
      "step": 213
    },
    {
      "epoch": 0.01909094964092957,
      "grad_norm": 140.57894897460938,
      "learning_rate": 3e-06,
      "loss": 26.7715,
      "step": 214
    },
    {
      "epoch": 0.019180159685980643,
      "grad_norm": 148.71348571777344,
      "learning_rate": 3e-06,
      "loss": 28.729,
      "step": 215
    },
    {
      "epoch": 0.019269369731031716,
      "grad_norm": 137.0448455810547,
      "learning_rate": 3e-06,
      "loss": 29.3048,
      "step": 216
    },
    {
      "completion_length": 241.0,
      "epoch": 0.019358579776082786,
      "grad_norm": 167.66650390625,
      "learning_rate": 3e-06,
      "loss": -43.2087,
      "reward": 1.6041667461395264,
      "reward_std": 0.9945478439331055,
      "rewards/correctness_reward_func": 1.375,
      "rewards/int_reward_func": 0.4270833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.1979166641831398,
      "step": 217,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.01944778982113386,
      "grad_norm": 168.0265350341797,
      "learning_rate": 3e-06,
      "loss": -45.6767,
      "step": 218
    },
    {
      "epoch": 0.019536999866184932,
      "grad_norm": 148.4340362548828,
      "learning_rate": 3e-06,
      "loss": -30.5571,
      "step": 219
    },
    {
      "epoch": 0.019626209911236005,
      "grad_norm": 139.6564178466797,
      "learning_rate": 3e-06,
      "loss": -44.7743,
      "step": 220
    },
    {
      "epoch": 0.01971541995628708,
      "grad_norm": 147.22129821777344,
      "learning_rate": 3e-06,
      "loss": -41.9365,
      "step": 221
    },
    {
      "epoch": 0.019804630001338152,
      "grad_norm": 190.81561279296875,
      "learning_rate": 3e-06,
      "loss": -48.2229,
      "step": 222
    },
    {
      "epoch": 0.019893840046389222,
      "grad_norm": 165.86917114257812,
      "learning_rate": 3e-06,
      "loss": -43.317,
      "step": 223
    },
    {
      "epoch": 0.019983050091440295,
      "grad_norm": 162.9475555419922,
      "learning_rate": 3e-06,
      "loss": -48.1878,
      "step": 224
    },
    {
      "epoch": 0.02007226013649137,
      "grad_norm": 179.08360290527344,
      "learning_rate": 3e-06,
      "loss": -33.3052,
      "step": 225
    },
    {
      "epoch": 0.02016147018154244,
      "grad_norm": 133.29290771484375,
      "learning_rate": 3e-06,
      "loss": -45.8993,
      "step": 226
    },
    {
      "epoch": 0.020250680226593515,
      "grad_norm": 155.86611938476562,
      "learning_rate": 3e-06,
      "loss": -43.9261,
      "step": 227
    },
    {
      "epoch": 0.020339890271644588,
      "grad_norm": 154.34974670410156,
      "learning_rate": 3e-06,
      "loss": -50.4381,
      "step": 228
    },
    {
      "completion_length": 239.43750762939453,
      "epoch": 0.02042910031669566,
      "grad_norm": 105.5196304321289,
      "learning_rate": 3e-06,
      "loss": 0.4059,
      "reward": 1.5316042304039001,
      "reward_std": 0.8583633303642273,
      "rewards/correctness_reward_func": 1.375,
      "rewards/int_reward_func": 0.3541666716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.19756250455975533,
      "step": 229,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.02051831036174673,
      "grad_norm": 119.5712890625,
      "learning_rate": 3e-06,
      "loss": -3.2594,
      "step": 230
    },
    {
      "epoch": 0.020607520406797804,
      "grad_norm": 128.1366424560547,
      "learning_rate": 3e-06,
      "loss": -12.0605,
      "step": 231
    },
    {
      "epoch": 0.020696730451848878,
      "grad_norm": 126.55559539794922,
      "learning_rate": 3e-06,
      "loss": -15.4799,
      "step": 232
    },
    {
      "epoch": 0.02078594049689995,
      "grad_norm": 137.93882751464844,
      "learning_rate": 3e-06,
      "loss": -18.5312,
      "step": 233
    },
    {
      "epoch": 0.020875150541951024,
      "grad_norm": 108.0162124633789,
      "learning_rate": 3e-06,
      "loss": -11.7573,
      "step": 234
    },
    {
      "epoch": 0.020964360587002098,
      "grad_norm": 118.95193481445312,
      "learning_rate": 3e-06,
      "loss": -1.1434,
      "step": 235
    },
    {
      "epoch": 0.02105357063205317,
      "grad_norm": 126.50416564941406,
      "learning_rate": 3e-06,
      "loss": -3.7423,
      "step": 236
    },
    {
      "epoch": 0.02114278067710424,
      "grad_norm": 130.68190002441406,
      "learning_rate": 3e-06,
      "loss": -14.5207,
      "step": 237
    },
    {
      "epoch": 0.021231990722155314,
      "grad_norm": 129.162109375,
      "learning_rate": 3e-06,
      "loss": -16.3237,
      "step": 238
    },
    {
      "epoch": 0.021321200767206387,
      "grad_norm": 145.95396423339844,
      "learning_rate": 3e-06,
      "loss": -20.6294,
      "step": 239
    },
    {
      "epoch": 0.02141041081225746,
      "grad_norm": 107.8385009765625,
      "learning_rate": 3e-06,
      "loss": -14.0773,
      "step": 240
    },
    {
      "completion_length": 206.9791717529297,
      "epoch": 0.021499620857308534,
      "grad_norm": 102.38019561767578,
      "learning_rate": 3e-06,
      "loss": -32.3486,
      "reward": 1.035479187965393,
      "reward_std": 0.7589404881000519,
      "rewards/correctness_reward_func": 0.8750000149011612,
      "rewards/int_reward_func": 0.3020833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.14160417020320892,
      "step": 241,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.021588830902359607,
      "grad_norm": 110.24679565429688,
      "learning_rate": 3e-06,
      "loss": -38.6199,
      "step": 242
    },
    {
      "epoch": 0.021678040947410677,
      "grad_norm": 118.22930145263672,
      "learning_rate": 3e-06,
      "loss": -52.9139,
      "step": 243
    },
    {
      "epoch": 0.02176725099246175,
      "grad_norm": 118.6080322265625,
      "learning_rate": 3e-06,
      "loss": -43.3805,
      "step": 244
    },
    {
      "epoch": 0.021856461037512823,
      "grad_norm": 106.9905776977539,
      "learning_rate": 3e-06,
      "loss": -36.7945,
      "step": 245
    },
    {
      "epoch": 0.021945671082563897,
      "grad_norm": 111.37010955810547,
      "learning_rate": 3e-06,
      "loss": -36.4452,
      "step": 246
    },
    {
      "epoch": 0.02203488112761497,
      "grad_norm": 104.93065643310547,
      "learning_rate": 3e-06,
      "loss": -34.2096,
      "step": 247
    },
    {
      "epoch": 0.022124091172666043,
      "grad_norm": 117.96737670898438,
      "learning_rate": 3e-06,
      "loss": -40.621,
      "step": 248
    },
    {
      "epoch": 0.022213301217717116,
      "grad_norm": 118.701904296875,
      "learning_rate": 3e-06,
      "loss": -54.4138,
      "step": 249
    },
    {
      "epoch": 0.022302511262768186,
      "grad_norm": 118.43307495117188,
      "learning_rate": 3e-06,
      "loss": -45.0393,
      "step": 250
    },
    {
      "epoch": 0.02239172130781926,
      "grad_norm": 114.41901397705078,
      "learning_rate": 3e-06,
      "loss": -37.6304,
      "step": 251
    },
    {
      "epoch": 0.022480931352870333,
      "grad_norm": 123.03970336914062,
      "learning_rate": 3e-06,
      "loss": -39.0638,
      "step": 252
    },
    {
      "completion_length": 211.20833587646484,
      "epoch": 0.022570141397921406,
      "grad_norm": 139.677734375,
      "learning_rate": 3e-06,
      "loss": -27.6756,
      "reward": 1.5570417046546936,
      "reward_std": 1.1208258867263794,
      "rewards/correctness_reward_func": 1.2916666865348816,
      "rewards/int_reward_func": 0.40625,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.14087500423192978,
      "step": 253,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.02265935144297248,
      "grad_norm": 145.56021118164062,
      "learning_rate": 3e-06,
      "loss": -40.3289,
      "step": 254
    },
    {
      "epoch": 0.022748561488023553,
      "grad_norm": 138.8564453125,
      "learning_rate": 3e-06,
      "loss": -39.1766,
      "step": 255
    },
    {
      "epoch": 0.022837771533074626,
      "grad_norm": 229.50186157226562,
      "learning_rate": 3e-06,
      "loss": -43.9568,
      "step": 256
    },
    {
      "epoch": 0.022926981578125696,
      "grad_norm": 138.42791748046875,
      "learning_rate": 3e-06,
      "loss": -52.4297,
      "step": 257
    },
    {
      "epoch": 0.02301619162317677,
      "grad_norm": 147.58364868164062,
      "learning_rate": 3e-06,
      "loss": -53.5477,
      "step": 258
    },
    {
      "epoch": 0.023105401668227842,
      "grad_norm": 140.5048828125,
      "learning_rate": 3e-06,
      "loss": -28.1418,
      "step": 259
    },
    {
      "epoch": 0.023194611713278915,
      "grad_norm": 139.11508178710938,
      "learning_rate": 3e-06,
      "loss": -42.7612,
      "step": 260
    },
    {
      "epoch": 0.02328382175832999,
      "grad_norm": 146.18580627441406,
      "learning_rate": 3e-06,
      "loss": -39.909,
      "step": 261
    },
    {
      "epoch": 0.023373031803381062,
      "grad_norm": 264.3643493652344,
      "learning_rate": 3e-06,
      "loss": -46.2595,
      "step": 262
    },
    {
      "epoch": 0.023462241848432132,
      "grad_norm": 154.1084747314453,
      "learning_rate": 3e-06,
      "loss": -55.2424,
      "step": 263
    },
    {
      "epoch": 0.023551451893483205,
      "grad_norm": 156.28662109375,
      "learning_rate": 3e-06,
      "loss": -55.6531,
      "step": 264
    },
    {
      "completion_length": 227.14583587646484,
      "epoch": 0.02364066193853428,
      "grad_norm": 115.90379333496094,
      "learning_rate": 3e-06,
      "loss": -30.7492,
      "reward": 1.6066043376922607,
      "reward_std": 0.8875448107719421,
      "rewards/correctness_reward_func": 1.2916666865348816,
      "rewards/int_reward_func": 0.3958333283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.010416666977107525,
      "rewards/xmlcount_reward_func": -0.09131250530481339,
      "step": 265,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.02372987198358535,
      "grad_norm": 113.86587524414062,
      "learning_rate": 3e-06,
      "loss": -29.6989,
      "step": 266
    },
    {
      "epoch": 0.023819082028636425,
      "grad_norm": 110.4273681640625,
      "learning_rate": 3e-06,
      "loss": -32.0614,
      "step": 267
    },
    {
      "epoch": 0.023908292073687498,
      "grad_norm": 111.84119415283203,
      "learning_rate": 3e-06,
      "loss": -31.9073,
      "step": 268
    },
    {
      "epoch": 0.02399750211873857,
      "grad_norm": 103.93081665039062,
      "learning_rate": 3e-06,
      "loss": -22.3506,
      "step": 269
    },
    {
      "epoch": 0.02408671216378964,
      "grad_norm": 120.32383728027344,
      "learning_rate": 3e-06,
      "loss": -28.5629,
      "step": 270
    },
    {
      "epoch": 0.024175922208840715,
      "grad_norm": 124.92536163330078,
      "learning_rate": 3e-06,
      "loss": -33.0776,
      "step": 271
    },
    {
      "epoch": 0.024265132253891788,
      "grad_norm": 119.54340362548828,
      "learning_rate": 3e-06,
      "loss": -31.6735,
      "step": 272
    },
    {
      "epoch": 0.02435434229894286,
      "grad_norm": 128.8444061279297,
      "learning_rate": 3e-06,
      "loss": -33.8033,
      "step": 273
    },
    {
      "epoch": 0.024443552343993934,
      "grad_norm": 123.08969116210938,
      "learning_rate": 3e-06,
      "loss": -34.4538,
      "step": 274
    },
    {
      "epoch": 0.024532762389045008,
      "grad_norm": 111.98983001708984,
      "learning_rate": 3e-06,
      "loss": -24.6449,
      "step": 275
    },
    {
      "epoch": 0.02462197243409608,
      "grad_norm": 123.31842041015625,
      "learning_rate": 3e-06,
      "loss": -31.1417,
      "step": 276
    },
    {
      "completion_length": 211.20833587646484,
      "epoch": 0.02471118247914715,
      "grad_norm": 83.42295837402344,
      "learning_rate": 3e-06,
      "loss": -49.3896,
      "reward": 1.6565834283828735,
      "reward_std": 0.7390342950820923,
      "rewards/correctness_reward_func": 1.375,
      "rewards/int_reward_func": 0.4166666716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.13508333638310432,
      "step": 277,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.024800392524198224,
      "grad_norm": 78.70240783691406,
      "learning_rate": 3e-06,
      "loss": -60.2538,
      "step": 278
    },
    {
      "epoch": 0.024889602569249297,
      "grad_norm": 87.03772735595703,
      "learning_rate": 3e-06,
      "loss": -54.701,
      "step": 279
    },
    {
      "epoch": 0.02497881261430037,
      "grad_norm": 105.03215789794922,
      "learning_rate": 3e-06,
      "loss": -50.647,
      "step": 280
    },
    {
      "epoch": 0.025068022659351444,
      "grad_norm": 94.19722747802734,
      "learning_rate": 3e-06,
      "loss": -53.7356,
      "step": 281
    },
    {
      "epoch": 0.025157232704402517,
      "grad_norm": 71.46943664550781,
      "learning_rate": 3e-06,
      "loss": -54.5847,
      "step": 282
    },
    {
      "epoch": 0.025246442749453587,
      "grad_norm": 90.4788589477539,
      "learning_rate": 3e-06,
      "loss": -50.5539,
      "step": 283
    },
    {
      "epoch": 0.02533565279450466,
      "grad_norm": 74.81779479980469,
      "learning_rate": 3e-06,
      "loss": -61.3813,
      "step": 284
    },
    {
      "epoch": 0.025424862839555733,
      "grad_norm": 85.80409240722656,
      "learning_rate": 3e-06,
      "loss": -55.7379,
      "step": 285
    },
    {
      "epoch": 0.025514072884606807,
      "grad_norm": 135.24191284179688,
      "learning_rate": 3e-06,
      "loss": -52.1614,
      "step": 286
    },
    {
      "epoch": 0.02560328292965788,
      "grad_norm": 94.01042175292969,
      "learning_rate": 3e-06,
      "loss": -55.4857,
      "step": 287
    },
    {
      "epoch": 0.025692492974708953,
      "grad_norm": 72.32071685791016,
      "learning_rate": 3e-06,
      "loss": -56.3565,
      "step": 288
    },
    {
      "completion_length": 221.45833587646484,
      "epoch": 0.025781703019760027,
      "grad_norm": 225.52276611328125,
      "learning_rate": 3e-06,
      "loss": -67.7081,
      "reward": 1.960687518119812,
      "reward_std": 0.8211362063884735,
      "rewards/correctness_reward_func": 1.625,
      "rewards/int_reward_func": 0.4479166716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.11222916468977928,
      "step": 289,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.025870913064811096,
      "grad_norm": 239.94651794433594,
      "learning_rate": 3e-06,
      "loss": -66.0587,
      "step": 290
    },
    {
      "epoch": 0.02596012310986217,
      "grad_norm": 173.2037353515625,
      "learning_rate": 3e-06,
      "loss": -59.416,
      "step": 291
    },
    {
      "epoch": 0.026049333154913243,
      "grad_norm": 228.50621032714844,
      "learning_rate": 3e-06,
      "loss": -70.6059,
      "step": 292
    },
    {
      "epoch": 0.026138543199964316,
      "grad_norm": 213.36802673339844,
      "learning_rate": 3e-06,
      "loss": -68.8733,
      "step": 293
    },
    {
      "epoch": 0.02622775324501539,
      "grad_norm": 389.8759460449219,
      "learning_rate": 3e-06,
      "loss": -108.725,
      "step": 294
    },
    {
      "epoch": 0.026316963290066463,
      "grad_norm": 241.96009826660156,
      "learning_rate": 3e-06,
      "loss": -73.0107,
      "step": 295
    },
    {
      "epoch": 0.026406173335117536,
      "grad_norm": 282.705322265625,
      "learning_rate": 3e-06,
      "loss": -71.4601,
      "step": 296
    },
    {
      "epoch": 0.026495383380168606,
      "grad_norm": 182.99859619140625,
      "learning_rate": 3e-06,
      "loss": -62.8503,
      "step": 297
    },
    {
      "epoch": 0.02658459342521968,
      "grad_norm": 237.8432159423828,
      "learning_rate": 3e-06,
      "loss": -76.095,
      "step": 298
    },
    {
      "epoch": 0.026673803470270752,
      "grad_norm": 224.10140991210938,
      "learning_rate": 3e-06,
      "loss": -71.9696,
      "step": 299
    },
    {
      "epoch": 0.026763013515321826,
      "grad_norm": 401.25421142578125,
      "learning_rate": 3e-06,
      "loss": -119.7468,
      "step": 300
    },
    {
      "completion_length": 181.00000762939453,
      "epoch": 0.0268522235603729,
      "grad_norm": 102.42095184326172,
      "learning_rate": 3e-06,
      "loss": 11.7663,
      "reward": 1.6927291750907898,
      "reward_std": 0.8399400115013123,
      "rewards/correctness_reward_func": 1.3333333730697632,
      "rewards/int_reward_func": 0.40625,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.0468541644513607,
      "step": 301,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.026941433605423972,
      "grad_norm": 97.59688568115234,
      "learning_rate": 3e-06,
      "loss": 4.3671,
      "step": 302
    },
    {
      "epoch": 0.027030643650475042,
      "grad_norm": 119.19691467285156,
      "learning_rate": 3e-06,
      "loss": 3.0122,
      "step": 303
    },
    {
      "epoch": 0.027119853695526115,
      "grad_norm": 102.54327392578125,
      "learning_rate": 3e-06,
      "loss": 5.7653,
      "step": 304
    },
    {
      "epoch": 0.02720906374057719,
      "grad_norm": 127.24678802490234,
      "learning_rate": 3e-06,
      "loss": 9.0808,
      "step": 305
    },
    {
      "epoch": 0.02729827378562826,
      "grad_norm": 115.35128784179688,
      "learning_rate": 3e-06,
      "loss": 9.7375,
      "step": 306
    },
    {
      "epoch": 0.027387483830679335,
      "grad_norm": 109.96597290039062,
      "learning_rate": 3e-06,
      "loss": 10.9794,
      "step": 307
    },
    {
      "epoch": 0.02747669387573041,
      "grad_norm": 116.67013549804688,
      "learning_rate": 3e-06,
      "loss": 4.0605,
      "step": 308
    },
    {
      "epoch": 0.02756590392078148,
      "grad_norm": 100.0082015991211,
      "learning_rate": 3e-06,
      "loss": 2.0719,
      "step": 309
    },
    {
      "epoch": 0.02765511396583255,
      "grad_norm": 103.2455062866211,
      "learning_rate": 3e-06,
      "loss": 3.753,
      "step": 310
    },
    {
      "epoch": 0.027744324010883625,
      "grad_norm": 139.74317932128906,
      "learning_rate": 3e-06,
      "loss": 7.435,
      "step": 311
    },
    {
      "epoch": 0.027833534055934698,
      "grad_norm": 126.05006408691406,
      "learning_rate": 3e-06,
      "loss": 8.1927,
      "step": 312
    },
    {
      "completion_length": 187.87500762939453,
      "epoch": 0.02792274410098577,
      "grad_norm": 182.04180908203125,
      "learning_rate": 3e-06,
      "loss": -90.2214,
      "reward": 1.3959375023841858,
      "reward_std": 0.7920421957969666,
      "rewards/correctness_reward_func": 1.0833333432674408,
      "rewards/int_reward_func": 0.40625,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.09364583343267441,
      "step": 313,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.028011954146036844,
      "grad_norm": 238.57090759277344,
      "learning_rate": 3e-06,
      "loss": -100.422,
      "step": 314
    },
    {
      "epoch": 0.028101164191087918,
      "grad_norm": 235.39544677734375,
      "learning_rate": 3e-06,
      "loss": -102.3354,
      "step": 315
    },
    {
      "epoch": 0.028190374236138988,
      "grad_norm": 223.8190460205078,
      "learning_rate": 3e-06,
      "loss": -109.9957,
      "step": 316
    },
    {
      "epoch": 0.02827958428119006,
      "grad_norm": 225.00672912597656,
      "learning_rate": 3e-06,
      "loss": -109.9375,
      "step": 317
    },
    {
      "epoch": 0.028368794326241134,
      "grad_norm": 247.57774353027344,
      "learning_rate": 3e-06,
      "loss": -125.4302,
      "step": 318
    },
    {
      "epoch": 0.028458004371292207,
      "grad_norm": 193.24212646484375,
      "learning_rate": 3e-06,
      "loss": -93.1797,
      "step": 319
    },
    {
      "epoch": 0.02854721441634328,
      "grad_norm": 264.4795227050781,
      "learning_rate": 3e-06,
      "loss": -104.7149,
      "step": 320
    },
    {
      "epoch": 0.028636424461394354,
      "grad_norm": 226.05810546875,
      "learning_rate": 3e-06,
      "loss": -107.5763,
      "step": 321
    },
    {
      "epoch": 0.028725634506445427,
      "grad_norm": 239.6378173828125,
      "learning_rate": 3e-06,
      "loss": -115.9954,
      "step": 322
    },
    {
      "epoch": 0.028814844551496497,
      "grad_norm": 240.8443145751953,
      "learning_rate": 3e-06,
      "loss": -117.6999,
      "step": 323
    },
    {
      "epoch": 0.02890405459654757,
      "grad_norm": 261.65643310546875,
      "learning_rate": 3e-06,
      "loss": -132.8027,
      "step": 324
    },
    {
      "completion_length": 189.8541717529297,
      "epoch": 0.028993264641598643,
      "grad_norm": 168.10452270507812,
      "learning_rate": 3e-06,
      "loss": 12.7588,
      "reward": 1.7772499918937683,
      "reward_std": 0.9346717596054077,
      "rewards/correctness_reward_func": 1.375,
      "rewards/int_reward_func": 0.4479166567325592,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.045666664838790894,
      "step": 325,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.029082474686649717,
      "grad_norm": 182.863037109375,
      "learning_rate": 3e-06,
      "loss": 0.3851,
      "step": 326
    },
    {
      "epoch": 0.02917168473170079,
      "grad_norm": 214.54574584960938,
      "learning_rate": 3e-06,
      "loss": 17.6492,
      "step": 327
    },
    {
      "epoch": 0.029260894776751863,
      "grad_norm": 187.80931091308594,
      "learning_rate": 3e-06,
      "loss": 7.692,
      "step": 328
    },
    {
      "epoch": 0.029350104821802937,
      "grad_norm": 195.0843505859375,
      "learning_rate": 3e-06,
      "loss": 12.8044,
      "step": 329
    },
    {
      "epoch": 0.029439314866854006,
      "grad_norm": 168.82028198242188,
      "learning_rate": 3e-06,
      "loss": -7.5147,
      "step": 330
    },
    {
      "epoch": 0.02952852491190508,
      "grad_norm": 176.14859008789062,
      "learning_rate": 3e-06,
      "loss": 10.9965,
      "step": 331
    },
    {
      "epoch": 0.029617734956956153,
      "grad_norm": 202.02247619628906,
      "learning_rate": 3e-06,
      "loss": -1.8138,
      "step": 332
    },
    {
      "epoch": 0.029706945002007226,
      "grad_norm": 216.37252807617188,
      "learning_rate": 3e-06,
      "loss": 15.5179,
      "step": 333
    },
    {
      "epoch": 0.0297961550470583,
      "grad_norm": 200.23558044433594,
      "learning_rate": 3e-06,
      "loss": 5.0549,
      "step": 334
    },
    {
      "epoch": 0.029885365092109373,
      "grad_norm": 177.7020263671875,
      "learning_rate": 3e-06,
      "loss": 11.199,
      "step": 335
    },
    {
      "epoch": 0.029974575137160443,
      "grad_norm": 170.23106384277344,
      "learning_rate": 3e-06,
      "loss": -10.9367,
      "step": 336
    },
    {
      "completion_length": 224.9166717529297,
      "epoch": 0.030063785182211516,
      "grad_norm": 127.8658218383789,
      "learning_rate": 3e-06,
      "loss": -59.3587,
      "reward": 1.6053959131240845,
      "reward_std": 0.5731277614831924,
      "rewards/correctness_reward_func": 1.3333333134651184,
      "rewards/int_reward_func": 0.4270833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.1550208330154419,
      "step": 337,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.03015299522726259,
      "grad_norm": 133.41494750976562,
      "learning_rate": 3e-06,
      "loss": -53.4241,
      "step": 338
    },
    {
      "epoch": 0.030242205272313662,
      "grad_norm": 170.7308807373047,
      "learning_rate": 3e-06,
      "loss": -65.4722,
      "step": 339
    },
    {
      "epoch": 0.030331415317364736,
      "grad_norm": 172.28118896484375,
      "learning_rate": 3e-06,
      "loss": -53.201,
      "step": 340
    },
    {
      "epoch": 0.03042062536241581,
      "grad_norm": 118.70462799072266,
      "learning_rate": 3e-06,
      "loss": -50.8363,
      "step": 341
    },
    {
      "epoch": 0.030509835407466882,
      "grad_norm": 143.119384765625,
      "learning_rate": 3e-06,
      "loss": -60.764,
      "step": 342
    },
    {
      "epoch": 0.030599045452517952,
      "grad_norm": 143.7277374267578,
      "learning_rate": 3e-06,
      "loss": -62.8186,
      "step": 343
    },
    {
      "epoch": 0.030688255497569025,
      "grad_norm": 157.5625,
      "learning_rate": 3e-06,
      "loss": -57.4741,
      "step": 344
    },
    {
      "epoch": 0.0307774655426201,
      "grad_norm": 191.64804077148438,
      "learning_rate": 3e-06,
      "loss": -71.2662,
      "step": 345
    },
    {
      "epoch": 0.030866675587671172,
      "grad_norm": 206.0039520263672,
      "learning_rate": 3e-06,
      "loss": -56.9883,
      "step": 346
    },
    {
      "epoch": 0.030955885632722245,
      "grad_norm": 132.1703643798828,
      "learning_rate": 3e-06,
      "loss": -54.0822,
      "step": 347
    },
    {
      "epoch": 0.03104509567777332,
      "grad_norm": 144.338623046875,
      "learning_rate": 3e-06,
      "loss": -66.1545,
      "step": 348
    },
    {
      "completion_length": 161.1041717529297,
      "epoch": 0.03113430572282439,
      "grad_norm": 176.29396057128906,
      "learning_rate": 3e-06,
      "loss": 2.568,
      "reward": 1.8858751058578491,
      "reward_std": 0.5198497474193573,
      "rewards/correctness_reward_func": 1.4583333134651184,
      "rewards/int_reward_func": 0.40625,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.021291667595505714,
      "step": 349,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.03122351576787546,
      "grad_norm": 162.1043701171875,
      "learning_rate": 3e-06,
      "loss": 11.6461,
      "step": 350
    },
    {
      "epoch": 0.03131272581292654,
      "grad_norm": 147.42918395996094,
      "learning_rate": 3e-06,
      "loss": 19.3425,
      "step": 351
    },
    {
      "epoch": 0.03140193585797761,
      "grad_norm": 146.65992736816406,
      "learning_rate": 3e-06,
      "loss": 4.1848,
      "step": 352
    },
    {
      "epoch": 0.03149114590302868,
      "grad_norm": 137.17474365234375,
      "learning_rate": 3e-06,
      "loss": 9.5202,
      "step": 353
    },
    {
      "epoch": 0.03158035594807975,
      "grad_norm": 176.24244689941406,
      "learning_rate": 3e-06,
      "loss": 24.3653,
      "step": 354
    },
    {
      "epoch": 0.031669565993130824,
      "grad_norm": 176.59144592285156,
      "learning_rate": 3e-06,
      "loss": 1.4274,
      "step": 355
    },
    {
      "epoch": 0.0317587760381819,
      "grad_norm": 161.0966339111328,
      "learning_rate": 3e-06,
      "loss": 10.7367,
      "step": 356
    },
    {
      "epoch": 0.03184798608323297,
      "grad_norm": 164.76675415039062,
      "learning_rate": 3e-06,
      "loss": 18.5651,
      "step": 357
    },
    {
      "epoch": 0.031937196128284044,
      "grad_norm": 227.15631103515625,
      "learning_rate": 3e-06,
      "loss": 2.7137,
      "step": 358
    },
    {
      "epoch": 0.03202640617333512,
      "grad_norm": 152.36514282226562,
      "learning_rate": 3e-06,
      "loss": 8.0471,
      "step": 359
    },
    {
      "epoch": 0.03211561621838619,
      "grad_norm": 162.82603454589844,
      "learning_rate": 3e-06,
      "loss": 24.2858,
      "step": 360
    },
    {
      "completion_length": 173.6666717529297,
      "epoch": 0.032204826263437264,
      "grad_norm": 112.11152648925781,
      "learning_rate": 3e-06,
      "loss": -47.6777,
      "reward": 1.9486668109893799,
      "reward_std": 0.614804282784462,
      "rewards/correctness_reward_func": 1.5,
      "rewards/int_reward_func": 0.4895833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": -0.04091667756438255,
      "step": 361,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.03229403630848834,
      "grad_norm": 212.41893005371094,
      "learning_rate": 3e-06,
      "loss": -57.8799,
      "step": 362
    },
    {
      "epoch": 0.03238324635353941,
      "grad_norm": 129.3856201171875,
      "learning_rate": 3e-06,
      "loss": -42.8209,
      "step": 363
    },
    {
      "epoch": 0.032472456398590484,
      "grad_norm": 90.02410888671875,
      "learning_rate": 3e-06,
      "loss": -37.0157,
      "step": 364
    },
    {
      "epoch": 0.03256166644364156,
      "grad_norm": 122.2002944946289,
      "learning_rate": 3e-06,
      "loss": -42.0898,
      "step": 365
    },
    {
      "epoch": 0.03265087648869262,
      "grad_norm": 114.75045776367188,
      "learning_rate": 3e-06,
      "loss": -37.9465,
      "step": 366
    },
    {
      "epoch": 0.0327400865337437,
      "grad_norm": 117.86136627197266,
      "learning_rate": 3e-06,
      "loss": -50.0162,
      "step": 367
    },
    {
      "epoch": 0.03282929657879477,
      "grad_norm": 224.55755615234375,
      "learning_rate": 3e-06,
      "loss": -62.4077,
      "step": 368
    },
    {
      "epoch": 0.03291850662384584,
      "grad_norm": 145.33380126953125,
      "learning_rate": 3e-06,
      "loss": -45.7888,
      "step": 369
    },
    {
      "epoch": 0.033007716668896916,
      "grad_norm": 107.85284423828125,
      "learning_rate": 3e-06,
      "loss": -38.8828,
      "step": 370
    },
    {
      "epoch": 0.03309692671394799,
      "grad_norm": 143.64854431152344,
      "learning_rate": 3e-06,
      "loss": -44.4827,
      "step": 371
    },
    {
      "epoch": 0.03318613675899906,
      "grad_norm": 120.4244155883789,
      "learning_rate": 3e-06,
      "loss": -40.0586,
      "step": 372
    },
    {
      "completion_length": 170.3125,
      "epoch": 0.033275346804050136,
      "grad_norm": 199.8765869140625,
      "learning_rate": 3e-06,
      "loss": -16.0424,
      "reward": 1.5020000338554382,
      "reward_std": 0.6375356912612915,
      "rewards/correctness_reward_func": 1.0416666865348816,
      "rewards/int_reward_func": 0.4479166716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.012416664510965347,
      "step": 373,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.03336455684910121,
      "grad_norm": 186.22103881835938,
      "learning_rate": 3e-06,
      "loss": -12.8464,
      "step": 374
    },
    {
      "epoch": 0.03345376689415228,
      "grad_norm": 133.84971618652344,
      "learning_rate": 3e-06,
      "loss": -16.0581,
      "step": 375
    },
    {
      "epoch": 0.033542976939203356,
      "grad_norm": 124.62361145019531,
      "learning_rate": 3e-06,
      "loss": -20.0662,
      "step": 376
    },
    {
      "epoch": 0.03363218698425443,
      "grad_norm": 177.62574768066406,
      "learning_rate": 3e-06,
      "loss": -16.7656,
      "step": 377
    },
    {
      "epoch": 0.0337213970293055,
      "grad_norm": 162.92381286621094,
      "learning_rate": 3e-06,
      "loss": -19.1936,
      "step": 378
    },
    {
      "epoch": 0.033810607074356576,
      "grad_norm": 167.49449157714844,
      "learning_rate": 3e-06,
      "loss": -16.9396,
      "step": 379
    },
    {
      "epoch": 0.03389981711940764,
      "grad_norm": 180.7197723388672,
      "learning_rate": 3e-06,
      "loss": -14.2949,
      "step": 380
    },
    {
      "epoch": 0.033989027164458716,
      "grad_norm": 158.6161346435547,
      "learning_rate": 3e-06,
      "loss": -18.1083,
      "step": 381
    },
    {
      "epoch": 0.03407823720950979,
      "grad_norm": 136.7860870361328,
      "learning_rate": 3e-06,
      "loss": -21.9306,
      "step": 382
    },
    {
      "epoch": 0.03416744725456086,
      "grad_norm": 200.51185607910156,
      "learning_rate": 3e-06,
      "loss": -18.8755,
      "step": 383
    },
    {
      "epoch": 0.034256657299611935,
      "grad_norm": 174.00477600097656,
      "learning_rate": 3e-06,
      "loss": -21.6131,
      "step": 384
    },
    {
      "completion_length": 168.0416717529297,
      "epoch": 0.03434586734466301,
      "grad_norm": 270.6558837890625,
      "learning_rate": 3e-06,
      "loss": -108.9543,
      "reward": 1.5282500386238098,
      "reward_std": 0.8121029734611511,
      "rewards/correctness_reward_func": 1.0416666567325592,
      "rewards/int_reward_func": 0.4479166716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.03866666788235307,
      "step": 385,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.03443507738971408,
      "grad_norm": 316.0130615234375,
      "learning_rate": 3e-06,
      "loss": -121.254,
      "step": 386
    },
    {
      "epoch": 0.034524287434765155,
      "grad_norm": 257.7696228027344,
      "learning_rate": 3e-06,
      "loss": -102.8911,
      "step": 387
    },
    {
      "epoch": 0.03461349747981623,
      "grad_norm": 305.755126953125,
      "learning_rate": 3e-06,
      "loss": -114.5659,
      "step": 388
    },
    {
      "epoch": 0.0347027075248673,
      "grad_norm": 219.5818328857422,
      "learning_rate": 3e-06,
      "loss": -114.7208,
      "step": 389
    },
    {
      "epoch": 0.034791917569918375,
      "grad_norm": 255.7522430419922,
      "learning_rate": 3e-06,
      "loss": -109.7934,
      "step": 390
    },
    {
      "epoch": 0.03488112761496945,
      "grad_norm": 291.77642822265625,
      "learning_rate": 3e-06,
      "loss": -116.5826,
      "step": 391
    },
    {
      "epoch": 0.03497033766002052,
      "grad_norm": 333.5157165527344,
      "learning_rate": 3e-06,
      "loss": -132.3897,
      "step": 392
    },
    {
      "epoch": 0.03505954770507159,
      "grad_norm": 267.8763122558594,
      "learning_rate": 3e-06,
      "loss": -110.8159,
      "step": 393
    },
    {
      "epoch": 0.03514875775012266,
      "grad_norm": 312.3733215332031,
      "learning_rate": 3e-06,
      "loss": -127.8078,
      "step": 394
    },
    {
      "epoch": 0.035237967795173734,
      "grad_norm": 242.0186309814453,
      "learning_rate": 3e-06,
      "loss": -123.3703,
      "step": 395
    },
    {
      "epoch": 0.03532717784022481,
      "grad_norm": 297.62847900390625,
      "learning_rate": 3e-06,
      "loss": -120.3945,
      "step": 396
    },
    {
      "completion_length": 171.45833587646484,
      "epoch": 0.03541638788527588,
      "grad_norm": 192.1062774658203,
      "learning_rate": 3e-06,
      "loss": -81.6189,
      "reward": 1.9622292518615723,
      "reward_std": 0.8264816105365753,
      "rewards/correctness_reward_func": 1.4583333134651184,
      "rewards/int_reward_func": 0.46875,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.0351458303630352,
      "step": 397,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.035505597930326954,
      "grad_norm": 202.8217010498047,
      "learning_rate": 3e-06,
      "loss": -67.341,
      "step": 398
    },
    {
      "epoch": 0.03559480797537803,
      "grad_norm": 218.82249450683594,
      "learning_rate": 3e-06,
      "loss": -65.3347,
      "step": 399
    },
    {
      "epoch": 0.0356840180204291,
      "grad_norm": 232.65196228027344,
      "learning_rate": 3e-06,
      "loss": -76.8989,
      "step": 400
    },
    {
      "epoch": 0.035773228065480174,
      "grad_norm": 198.04103088378906,
      "learning_rate": 3e-06,
      "loss": -74.488,
      "step": 401
    },
    {
      "epoch": 0.03586243811053125,
      "grad_norm": 211.86273193359375,
      "learning_rate": 3e-06,
      "loss": -75.6096,
      "step": 402
    },
    {
      "epoch": 0.03595164815558232,
      "grad_norm": 218.77589416503906,
      "learning_rate": 3e-06,
      "loss": -87.9174,
      "step": 403
    },
    {
      "epoch": 0.036040858200633394,
      "grad_norm": 243.4962615966797,
      "learning_rate": 3e-06,
      "loss": -70.1026,
      "step": 404
    },
    {
      "epoch": 0.03613006824568447,
      "grad_norm": 242.30494689941406,
      "learning_rate": 3e-06,
      "loss": -71.4579,
      "step": 405
    },
    {
      "epoch": 0.03621927829073553,
      "grad_norm": 274.28948974609375,
      "learning_rate": 3e-06,
      "loss": -83.8044,
      "step": 406
    },
    {
      "epoch": 0.03630848833578661,
      "grad_norm": 257.0942077636719,
      "learning_rate": 3e-06,
      "loss": -82.445,
      "step": 407
    },
    {
      "epoch": 0.03639769838083768,
      "grad_norm": 255.2320556640625,
      "learning_rate": 3e-06,
      "loss": -81.9402,
      "step": 408
    },
    {
      "completion_length": 125.70833587646484,
      "epoch": 0.03648690842588875,
      "grad_norm": 185.7193145751953,
      "learning_rate": 3e-06,
      "loss": 46.4742,
      "reward": 1.8683959245681763,
      "reward_std": 0.8172085583209991,
      "rewards/correctness_reward_func": 1.3333333730697632,
      "rewards/int_reward_func": 0.4270833432674408,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.10797916725277901,
      "step": 409,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.03657611847093983,
      "grad_norm": 218.23338317871094,
      "learning_rate": 3e-06,
      "loss": 29.3517,
      "step": 410
    },
    {
      "epoch": 0.0366653285159909,
      "grad_norm": 180.90330505371094,
      "learning_rate": 3e-06,
      "loss": 55.3162,
      "step": 411
    },
    {
      "epoch": 0.03675453856104197,
      "grad_norm": 216.37953186035156,
      "learning_rate": 3e-06,
      "loss": 50.2096,
      "step": 412
    },
    {
      "epoch": 0.036843748606093046,
      "grad_norm": 198.8724822998047,
      "learning_rate": 3e-06,
      "loss": 51.636,
      "step": 413
    },
    {
      "epoch": 0.03693295865114412,
      "grad_norm": 184.89627075195312,
      "learning_rate": 3e-06,
      "loss": 44.8369,
      "step": 414
    },
    {
      "epoch": 0.03702216869619519,
      "grad_norm": 167.6713104248047,
      "learning_rate": 3e-06,
      "loss": 44.1546,
      "step": 415
    },
    {
      "epoch": 0.037111378741246266,
      "grad_norm": 192.13140869140625,
      "learning_rate": 3e-06,
      "loss": 27.4686,
      "step": 416
    },
    {
      "epoch": 0.03720058878629734,
      "grad_norm": 177.4408721923828,
      "learning_rate": 3e-06,
      "loss": 53.37,
      "step": 417
    },
    {
      "epoch": 0.03728979883134841,
      "grad_norm": 223.81668090820312,
      "learning_rate": 3e-06,
      "loss": 45.2759,
      "step": 418
    },
    {
      "epoch": 0.037379008876399486,
      "grad_norm": 207.5684356689453,
      "learning_rate": 3e-06,
      "loss": 46.924,
      "step": 419
    },
    {
      "epoch": 0.03746821892145055,
      "grad_norm": 180.81484985351562,
      "learning_rate": 3e-06,
      "loss": 42.0928,
      "step": 420
    },
    {
      "completion_length": 153.89583587646484,
      "epoch": 0.037557428966501626,
      "grad_norm": 247.00067138671875,
      "learning_rate": 3e-06,
      "loss": 14.4654,
      "reward": 1.776770830154419,
      "reward_std": 0.6972799003124237,
      "rewards/correctness_reward_func": 1.25,
      "rewards/int_reward_func": 0.4479166716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.07885416969656944,
      "step": 421,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.0376466390115527,
      "grad_norm": 244.66824340820312,
      "learning_rate": 3e-06,
      "loss": 34.9601,
      "step": 422
    },
    {
      "epoch": 0.03773584905660377,
      "grad_norm": 276.21539306640625,
      "learning_rate": 3e-06,
      "loss": 14.6261,
      "step": 423
    },
    {
      "epoch": 0.037825059101654845,
      "grad_norm": 288.96246337890625,
      "learning_rate": 3e-06,
      "loss": 41.9368,
      "step": 424
    },
    {
      "epoch": 0.03791426914670592,
      "grad_norm": 303.6945495605469,
      "learning_rate": 3e-06,
      "loss": 23.9119,
      "step": 425
    },
    {
      "epoch": 0.03800347919175699,
      "grad_norm": 274.27142333984375,
      "learning_rate": 3e-06,
      "loss": 27.399,
      "step": 426
    },
    {
      "epoch": 0.038092689236808065,
      "grad_norm": 233.245361328125,
      "learning_rate": 3e-06,
      "loss": 10.175,
      "step": 427
    },
    {
      "epoch": 0.03818189928185914,
      "grad_norm": 256.8597412109375,
      "learning_rate": 3e-06,
      "loss": 31.9908,
      "step": 428
    },
    {
      "epoch": 0.03827110932691021,
      "grad_norm": 270.4859619140625,
      "learning_rate": 3e-06,
      "loss": 10.7867,
      "step": 429
    },
    {
      "epoch": 0.038360319371961285,
      "grad_norm": 301.17181396484375,
      "learning_rate": 3e-06,
      "loss": 40.6524,
      "step": 430
    },
    {
      "epoch": 0.03844952941701236,
      "grad_norm": 303.94488525390625,
      "learning_rate": 3e-06,
      "loss": 21.1706,
      "step": 431
    },
    {
      "epoch": 0.03853873946206343,
      "grad_norm": 258.3034973144531,
      "learning_rate": 3e-06,
      "loss": 22.7622,
      "step": 432
    },
    {
      "completion_length": 107.54166793823242,
      "epoch": 0.0386279495071145,
      "grad_norm": 134.02154541015625,
      "learning_rate": 3e-06,
      "loss": -22.4999,
      "reward": 2.2951666712760925,
      "reward_std": 0.38126008585095406,
      "rewards/correctness_reward_func": 1.5833333134651184,
      "rewards/int_reward_func": 0.5,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.21183334290981293,
      "step": 433,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.03871715955216557,
      "grad_norm": 191.98023986816406,
      "learning_rate": 3e-06,
      "loss": -14.0492,
      "step": 434
    },
    {
      "epoch": 0.038806369597216644,
      "grad_norm": 154.30328369140625,
      "learning_rate": 3e-06,
      "loss": -10.9444,
      "step": 435
    },
    {
      "epoch": 0.03889557964226772,
      "grad_norm": 134.01214599609375,
      "learning_rate": 3e-06,
      "loss": -17.7574,
      "step": 436
    },
    {
      "epoch": 0.03898478968731879,
      "grad_norm": 132.3379364013672,
      "learning_rate": 3e-06,
      "loss": -18.487,
      "step": 437
    },
    {
      "epoch": 0.039073999732369864,
      "grad_norm": 146.31573486328125,
      "learning_rate": 3e-06,
      "loss": -12.7164,
      "step": 438
    },
    {
      "epoch": 0.03916320977742094,
      "grad_norm": 136.05592346191406,
      "learning_rate": 3e-06,
      "loss": -23.1439,
      "step": 439
    },
    {
      "epoch": 0.03925241982247201,
      "grad_norm": 138.1117706298828,
      "learning_rate": 3e-06,
      "loss": -15.7255,
      "step": 440
    },
    {
      "epoch": 0.039341629867523084,
      "grad_norm": 166.34922790527344,
      "learning_rate": 3e-06,
      "loss": -12.6375,
      "step": 441
    },
    {
      "epoch": 0.03943083991257416,
      "grad_norm": 132.994140625,
      "learning_rate": 3e-06,
      "loss": -20.1431,
      "step": 442
    },
    {
      "epoch": 0.03952004995762523,
      "grad_norm": 129.54771423339844,
      "learning_rate": 3e-06,
      "loss": -19.3478,
      "step": 443
    },
    {
      "epoch": 0.039609260002676304,
      "grad_norm": 152.91607666015625,
      "learning_rate": 3e-06,
      "loss": -14.3018,
      "step": 444
    },
    {
      "completion_length": 160.7916717529297,
      "epoch": 0.03969847004772738,
      "grad_norm": 165.9615020751953,
      "learning_rate": 3e-06,
      "loss": 38.8687,
      "reward": 1.6081042885780334,
      "reward_std": 0.45602357387542725,
      "rewards/correctness_reward_func": 1.0833333432674408,
      "rewards/int_reward_func": 0.4479166716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.0768541656434536,
      "step": 445,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.039787680092778444,
      "grad_norm": 165.71353149414062,
      "learning_rate": 3e-06,
      "loss": 32.8818,
      "step": 446
    },
    {
      "epoch": 0.03987689013782952,
      "grad_norm": 185.27174377441406,
      "learning_rate": 3e-06,
      "loss": 29.8206,
      "step": 447
    },
    {
      "epoch": 0.03996610018288059,
      "grad_norm": 163.79954528808594,
      "learning_rate": 3e-06,
      "loss": 28.2399,
      "step": 448
    },
    {
      "epoch": 0.04005531022793166,
      "grad_norm": 167.2331085205078,
      "learning_rate": 3e-06,
      "loss": 42.2504,
      "step": 449
    },
    {
      "epoch": 0.04014452027298274,
      "grad_norm": 157.44320678710938,
      "learning_rate": 3e-06,
      "loss": 47.0631,
      "step": 450
    },
    {
      "epoch": 0.04023373031803381,
      "grad_norm": 167.7976837158203,
      "learning_rate": 3e-06,
      "loss": 37.7299,
      "step": 451
    },
    {
      "epoch": 0.04032294036308488,
      "grad_norm": 171.96420288085938,
      "learning_rate": 3e-06,
      "loss": 31.7018,
      "step": 452
    },
    {
      "epoch": 0.040412150408135956,
      "grad_norm": 164.95046997070312,
      "learning_rate": 3e-06,
      "loss": 28.9306,
      "step": 453
    },
    {
      "epoch": 0.04050136045318703,
      "grad_norm": 146.903076171875,
      "learning_rate": 3e-06,
      "loss": 26.551,
      "step": 454
    },
    {
      "epoch": 0.0405905704982381,
      "grad_norm": 182.6881561279297,
      "learning_rate": 3e-06,
      "loss": 41.0788,
      "step": 455
    },
    {
      "epoch": 0.040679780543289176,
      "grad_norm": 147.5907440185547,
      "learning_rate": 3e-06,
      "loss": 44.0529,
      "step": 456
    },
    {
      "completion_length": 143.8541717529297,
      "epoch": 0.04076899058834025,
      "grad_norm": 252.46615600585938,
      "learning_rate": 3e-06,
      "loss": -60.641,
      "reward": 1.8890208005905151,
      "reward_std": 0.4024546667933464,
      "rewards/correctness_reward_func": 1.375,
      "rewards/int_reward_func": 0.4166666716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.09735416248440742,
      "step": 457,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.04085820063339132,
      "grad_norm": 223.07171630859375,
      "learning_rate": 3e-06,
      "loss": -42.0093,
      "step": 458
    },
    {
      "epoch": 0.040947410678442396,
      "grad_norm": 237.3083953857422,
      "learning_rate": 3e-06,
      "loss": -57.9657,
      "step": 459
    },
    {
      "epoch": 0.04103662072349346,
      "grad_norm": 225.29269409179688,
      "learning_rate": 3e-06,
      "loss": -41.8661,
      "step": 460
    },
    {
      "epoch": 0.041125830768544536,
      "grad_norm": 210.0297088623047,
      "learning_rate": 3e-06,
      "loss": -47.2305,
      "step": 461
    },
    {
      "epoch": 0.04121504081359561,
      "grad_norm": 262.04473876953125,
      "learning_rate": 3e-06,
      "loss": -50.9342,
      "step": 462
    },
    {
      "epoch": 0.04130425085864668,
      "grad_norm": 252.53802490234375,
      "learning_rate": 3e-06,
      "loss": -62.6208,
      "step": 463
    },
    {
      "epoch": 0.041393460903697755,
      "grad_norm": 221.40121459960938,
      "learning_rate": 3e-06,
      "loss": -43.7282,
      "step": 464
    },
    {
      "epoch": 0.04148267094874883,
      "grad_norm": 231.59335327148438,
      "learning_rate": 3e-06,
      "loss": -60.5224,
      "step": 465
    },
    {
      "epoch": 0.0415718809937999,
      "grad_norm": 207.73471069335938,
      "learning_rate": 3e-06,
      "loss": -44.5223,
      "step": 466
    },
    {
      "epoch": 0.041661091038850975,
      "grad_norm": 217.08779907226562,
      "learning_rate": 3e-06,
      "loss": -50.6021,
      "step": 467
    },
    {
      "epoch": 0.04175030108390205,
      "grad_norm": 254.29269409179688,
      "learning_rate": 3e-06,
      "loss": -55.4887,
      "step": 468
    },
    {
      "completion_length": 122.50000381469727,
      "epoch": 0.04183951112895312,
      "grad_norm": 147.28335571289062,
      "learning_rate": 3e-06,
      "loss": 12.9864,
      "reward": 2.310395896434784,
      "reward_std": 0.44204503297805786,
      "rewards/correctness_reward_func": 1.6666666269302368,
      "rewards/int_reward_func": 0.46875,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.010416666977107525,
      "rewards/xmlcount_reward_func": 0.16456249356269836,
      "step": 469,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.041928721174004195,
      "grad_norm": 161.9978485107422,
      "learning_rate": 3e-06,
      "loss": 34.7546,
      "step": 470
    },
    {
      "epoch": 0.04201793121905527,
      "grad_norm": 165.3116455078125,
      "learning_rate": 3e-06,
      "loss": 17.4188,
      "step": 471
    },
    {
      "epoch": 0.04210714126410634,
      "grad_norm": 142.81861877441406,
      "learning_rate": 3e-06,
      "loss": 18.3006,
      "step": 472
    },
    {
      "epoch": 0.04219635130915741,
      "grad_norm": 168.01116943359375,
      "learning_rate": 3e-06,
      "loss": 17.6413,
      "step": 473
    },
    {
      "epoch": 0.04228556135420848,
      "grad_norm": 207.03326416015625,
      "learning_rate": 3e-06,
      "loss": 15.9259,
      "step": 474
    },
    {
      "epoch": 0.042374771399259555,
      "grad_norm": 141.62599182128906,
      "learning_rate": 3e-06,
      "loss": 12.4355,
      "step": 475
    },
    {
      "epoch": 0.04246398144431063,
      "grad_norm": 180.9537353515625,
      "learning_rate": 3e-06,
      "loss": 32.9299,
      "step": 476
    },
    {
      "epoch": 0.0425531914893617,
      "grad_norm": 163.92254638671875,
      "learning_rate": 3e-06,
      "loss": 17.1089,
      "step": 477
    },
    {
      "epoch": 0.042642401534412774,
      "grad_norm": 145.9250030517578,
      "learning_rate": 3e-06,
      "loss": 17.069,
      "step": 478
    },
    {
      "epoch": 0.04273161157946385,
      "grad_norm": 152.68319702148438,
      "learning_rate": 3e-06,
      "loss": 16.6025,
      "step": 479
    },
    {
      "epoch": 0.04282082162451492,
      "grad_norm": 222.65997314453125,
      "learning_rate": 3e-06,
      "loss": 14.2435,
      "step": 480
    },
    {
      "completion_length": 116.66666793823242,
      "epoch": 0.042910031669565994,
      "grad_norm": 166.38722229003906,
      "learning_rate": 3e-06,
      "loss": 13.4616,
      "reward": 1.5636458992958069,
      "reward_std": 0.6559399664402008,
      "rewards/correctness_reward_func": 0.9999999701976776,
      "rewards/int_reward_func": 0.4166666716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.1469791643321514,
      "step": 481,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.04299924171461707,
      "grad_norm": 222.22879028320312,
      "learning_rate": 3e-06,
      "loss": 18.8121,
      "step": 482
    },
    {
      "epoch": 0.04308845175966814,
      "grad_norm": 221.17059326171875,
      "learning_rate": 3e-06,
      "loss": 21.4877,
      "step": 483
    },
    {
      "epoch": 0.043177661804719214,
      "grad_norm": 142.53189086914062,
      "learning_rate": 3e-06,
      "loss": 16.8492,
      "step": 484
    },
    {
      "epoch": 0.04326687184977029,
      "grad_norm": 170.13198852539062,
      "learning_rate": 3e-06,
      "loss": 20.7898,
      "step": 485
    },
    {
      "epoch": 0.043356081894821354,
      "grad_norm": 161.23110961914062,
      "learning_rate": 3e-06,
      "loss": 12.8851,
      "step": 486
    },
    {
      "epoch": 0.04344529193987243,
      "grad_norm": 175.66587829589844,
      "learning_rate": 3e-06,
      "loss": 10.8873,
      "step": 487
    },
    {
      "epoch": 0.0435345019849235,
      "grad_norm": 195.75050354003906,
      "learning_rate": 3e-06,
      "loss": 15.6694,
      "step": 488
    },
    {
      "epoch": 0.04362371202997457,
      "grad_norm": 190.2042236328125,
      "learning_rate": 3e-06,
      "loss": 19.0926,
      "step": 489
    },
    {
      "epoch": 0.04371292207502565,
      "grad_norm": 146.10504150390625,
      "learning_rate": 3e-06,
      "loss": 13.8907,
      "step": 490
    },
    {
      "epoch": 0.04380213212007672,
      "grad_norm": 149.26614379882812,
      "learning_rate": 3e-06,
      "loss": 17.0683,
      "step": 491
    },
    {
      "epoch": 0.04389134216512779,
      "grad_norm": 178.4593963623047,
      "learning_rate": 3e-06,
      "loss": 10.2799,
      "step": 492
    },
    {
      "completion_length": 137.50000762939453,
      "epoch": 0.043980552210178867,
      "grad_norm": 87.93815612792969,
      "learning_rate": 3e-06,
      "loss": -35.7257,
      "reward": 2.5712709426879883,
      "reward_std": 0.18458116799592972,
      "rewards/correctness_reward_func": 1.9583333134651184,
      "rewards/int_reward_func": 0.5,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.11293749511241913,
      "step": 493,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.04406976225522994,
      "grad_norm": 78.96405029296875,
      "learning_rate": 3e-06,
      "loss": -37.1011,
      "step": 494
    },
    {
      "epoch": 0.04415897230028101,
      "grad_norm": 113.70293426513672,
      "learning_rate": 3e-06,
      "loss": -46.3151,
      "step": 495
    },
    {
      "epoch": 0.044248182345332086,
      "grad_norm": 90.45478820800781,
      "learning_rate": 3e-06,
      "loss": -43.5143,
      "step": 496
    },
    {
      "epoch": 0.04433739239038316,
      "grad_norm": 106.42904663085938,
      "learning_rate": 3e-06,
      "loss": -42.7944,
      "step": 497
    },
    {
      "epoch": 0.04442660243543423,
      "grad_norm": 106.20608520507812,
      "learning_rate": 3e-06,
      "loss": -48.1815,
      "step": 498
    },
    {
      "epoch": 0.0445158124804853,
      "grad_norm": 93.41876220703125,
      "learning_rate": 3e-06,
      "loss": -37.7992,
      "step": 499
    },
    {
      "epoch": 0.04460502252553637,
      "grad_norm": 91.40050506591797,
      "learning_rate": 3e-06,
      "loss": -38.651,
      "step": 500
    },
    {
      "epoch": 0.044694232570587446,
      "grad_norm": 116.4251480102539,
      "learning_rate": 3e-06,
      "loss": -49.2276,
      "step": 501
    },
    {
      "epoch": 0.04478344261563852,
      "grad_norm": 92.2903060913086,
      "learning_rate": 3e-06,
      "loss": -46.591,
      "step": 502
    },
    {
      "epoch": 0.04487265266068959,
      "grad_norm": 110.7293472290039,
      "learning_rate": 3e-06,
      "loss": -45.7648,
      "step": 503
    },
    {
      "epoch": 0.044961862705740666,
      "grad_norm": 108.03923797607422,
      "learning_rate": 3e-06,
      "loss": -51.9054,
      "step": 504
    },
    {
      "completion_length": 131.14583587646484,
      "epoch": 0.04505107275079174,
      "grad_norm": 452.2289733886719,
      "learning_rate": 3e-06,
      "loss": 29.1255,
      "reward": 1.9235833883285522,
      "reward_std": 0.813209742307663,
      "rewards/correctness_reward_func": 1.3333333730697632,
      "rewards/int_reward_func": 0.4270833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.16316666454076767,
      "step": 505,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.04514028279584281,
      "grad_norm": 284.8124694824219,
      "learning_rate": 3e-06,
      "loss": 31.7163,
      "step": 506
    },
    {
      "epoch": 0.045229492840893885,
      "grad_norm": 361.7442626953125,
      "learning_rate": 3e-06,
      "loss": 27.2282,
      "step": 507
    },
    {
      "epoch": 0.04531870288594496,
      "grad_norm": 283.879638671875,
      "learning_rate": 3e-06,
      "loss": 42.2817,
      "step": 508
    },
    {
      "epoch": 0.04540791293099603,
      "grad_norm": 316.19000244140625,
      "learning_rate": 3e-06,
      "loss": 26.7891,
      "step": 509
    },
    {
      "epoch": 0.045497122976047105,
      "grad_norm": 370.62652587890625,
      "learning_rate": 3e-06,
      "loss": 34.1636,
      "step": 510
    },
    {
      "epoch": 0.04558633302109818,
      "grad_norm": 273.1391296386719,
      "learning_rate": 3e-06,
      "loss": 27.6705,
      "step": 511
    },
    {
      "epoch": 0.04567554306614925,
      "grad_norm": 307.9808044433594,
      "learning_rate": 3e-06,
      "loss": 28.9577,
      "step": 512
    },
    {
      "epoch": 0.04576475311120032,
      "grad_norm": 374.3335876464844,
      "learning_rate": 3e-06,
      "loss": 22.2669,
      "step": 513
    },
    {
      "epoch": 0.04585396315625139,
      "grad_norm": 395.0052795410156,
      "learning_rate": 3e-06,
      "loss": 39.239,
      "step": 514
    },
    {
      "epoch": 0.045943173201302465,
      "grad_norm": 539.7128295898438,
      "learning_rate": 3e-06,
      "loss": 25.6475,
      "step": 515
    },
    {
      "epoch": 0.04603238324635354,
      "grad_norm": 348.81170654296875,
      "learning_rate": 3e-06,
      "loss": 31.7762,
      "step": 516
    },
    {
      "completion_length": 115.83333587646484,
      "epoch": 0.04612159329140461,
      "grad_norm": 177.44480895996094,
      "learning_rate": 3e-06,
      "loss": -44.7672,
      "reward": 2.343208432197571,
      "reward_std": 0.4359329864382744,
      "rewards/correctness_reward_func": 1.6666666269302368,
      "rewards/int_reward_func": 0.4895833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.18695833534002304,
      "step": 517,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.046210803336455684,
      "grad_norm": 199.65908813476562,
      "learning_rate": 3e-06,
      "loss": -51.4622,
      "step": 518
    },
    {
      "epoch": 0.04630001338150676,
      "grad_norm": 175.2034149169922,
      "learning_rate": 3e-06,
      "loss": -55.1536,
      "step": 519
    },
    {
      "epoch": 0.04638922342655783,
      "grad_norm": 160.91688537597656,
      "learning_rate": 3e-06,
      "loss": -37.3164,
      "step": 520
    },
    {
      "epoch": 0.046478433471608904,
      "grad_norm": 165.5592498779297,
      "learning_rate": 3e-06,
      "loss": -42.9147,
      "step": 521
    },
    {
      "epoch": 0.04656764351665998,
      "grad_norm": 154.5955047607422,
      "learning_rate": 3e-06,
      "loss": -48.9731,
      "step": 522
    },
    {
      "epoch": 0.04665685356171105,
      "grad_norm": 202.06838989257812,
      "learning_rate": 3e-06,
      "loss": -46.1992,
      "step": 523
    },
    {
      "epoch": 0.046746063606762124,
      "grad_norm": 216.6766357421875,
      "learning_rate": 3e-06,
      "loss": -54.2271,
      "step": 524
    },
    {
      "epoch": 0.0468352736518132,
      "grad_norm": 212.4103240966797,
      "learning_rate": 3e-06,
      "loss": -59.7351,
      "step": 525
    },
    {
      "epoch": 0.046924483696864264,
      "grad_norm": 160.86546325683594,
      "learning_rate": 3e-06,
      "loss": -40.5866,
      "step": 526
    },
    {
      "epoch": 0.04701369374191534,
      "grad_norm": 171.24478149414062,
      "learning_rate": 3e-06,
      "loss": -47.2665,
      "step": 527
    },
    {
      "epoch": 0.04710290378696641,
      "grad_norm": 165.52357482910156,
      "learning_rate": 3e-06,
      "loss": -53.477,
      "step": 528
    },
    {
      "completion_length": 119.77083587646484,
      "epoch": 0.047192113832017483,
      "grad_norm": 382.3139953613281,
      "learning_rate": 3e-06,
      "loss": 86.4678,
      "reward": 1.9181458950042725,
      "reward_std": 0.6776820421218872,
      "rewards/correctness_reward_func": 1.3333333134651184,
      "rewards/int_reward_func": 0.4166666716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.1681458279490471,
      "step": 529,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.04728132387706856,
      "grad_norm": 264.3744812011719,
      "learning_rate": 3e-06,
      "loss": 87.2703,
      "step": 530
    },
    {
      "epoch": 0.04737053392211963,
      "grad_norm": 273.3477783203125,
      "learning_rate": 3e-06,
      "loss": 74.9615,
      "step": 531
    },
    {
      "epoch": 0.0474597439671707,
      "grad_norm": 326.87078857421875,
      "learning_rate": 3e-06,
      "loss": 90.7838,
      "step": 532
    },
    {
      "epoch": 0.04754895401222178,
      "grad_norm": 294.74041748046875,
      "learning_rate": 3e-06,
      "loss": 102.1572,
      "step": 533
    },
    {
      "epoch": 0.04763816405727285,
      "grad_norm": 312.48626708984375,
      "learning_rate": 3e-06,
      "loss": 93.4423,
      "step": 534
    },
    {
      "epoch": 0.04772737410232392,
      "grad_norm": 383.2833557128906,
      "learning_rate": 3e-06,
      "loss": 86.6238,
      "step": 535
    },
    {
      "epoch": 0.047816584147374996,
      "grad_norm": 316.42926025390625,
      "learning_rate": 3e-06,
      "loss": 86.6783,
      "step": 536
    },
    {
      "epoch": 0.04790579419242607,
      "grad_norm": 268.37506103515625,
      "learning_rate": 3e-06,
      "loss": 72.0206,
      "step": 537
    },
    {
      "epoch": 0.04799500423747714,
      "grad_norm": 337.4726867675781,
      "learning_rate": 3e-06,
      "loss": 87.3427,
      "step": 538
    },
    {
      "epoch": 0.04808421428252821,
      "grad_norm": 284.79827880859375,
      "learning_rate": 3e-06,
      "loss": 98.66,
      "step": 539
    },
    {
      "epoch": 0.04817342432757928,
      "grad_norm": 321.74609375,
      "learning_rate": 3e-06,
      "loss": 87.6142,
      "step": 540
    },
    {
      "completion_length": 125.37500762939453,
      "epoch": 0.048262634372630356,
      "grad_norm": 187.67727661132812,
      "learning_rate": 3e-06,
      "loss": -1.5514,
      "reward": 1.9883333444595337,
      "reward_std": 0.6124217808246613,
      "rewards/correctness_reward_func": 1.4166666865348816,
      "rewards/int_reward_func": 0.4375,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.1341666616499424,
      "step": 541,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.04835184441768143,
      "grad_norm": 202.66070556640625,
      "learning_rate": 3e-06,
      "loss": 1.9568,
      "step": 542
    },
    {
      "epoch": 0.0484410544627325,
      "grad_norm": 180.55126953125,
      "learning_rate": 3e-06,
      "loss": -6.8013,
      "step": 543
    },
    {
      "epoch": 0.048530264507783576,
      "grad_norm": 161.08514404296875,
      "learning_rate": 3e-06,
      "loss": -3.4725,
      "step": 544
    },
    {
      "epoch": 0.04861947455283465,
      "grad_norm": 220.28076171875,
      "learning_rate": 3e-06,
      "loss": 0.0972,
      "step": 545
    },
    {
      "epoch": 0.04870868459788572,
      "grad_norm": 321.00994873046875,
      "learning_rate": 3e-06,
      "loss": -10.5751,
      "step": 546
    },
    {
      "epoch": 0.048797894642936795,
      "grad_norm": 197.3623046875,
      "learning_rate": 3e-06,
      "loss": -2.6929,
      "step": 547
    },
    {
      "epoch": 0.04888710468798787,
      "grad_norm": 213.94691467285156,
      "learning_rate": 3e-06,
      "loss": 1.3407,
      "step": 548
    },
    {
      "epoch": 0.04897631473303894,
      "grad_norm": 254.2111053466797,
      "learning_rate": 3e-06,
      "loss": -7.3544,
      "step": 549
    },
    {
      "epoch": 0.049065524778090015,
      "grad_norm": 155.93460083007812,
      "learning_rate": 3e-06,
      "loss": -5.7191,
      "step": 550
    },
    {
      "epoch": 0.04915473482314109,
      "grad_norm": 175.17147827148438,
      "learning_rate": 3e-06,
      "loss": -2.0149,
      "step": 551
    },
    {
      "epoch": 0.04924394486819216,
      "grad_norm": 220.0244140625,
      "learning_rate": 3e-06,
      "loss": -12.1544,
      "step": 552
    },
    {
      "completion_length": 138.62500762939453,
      "epoch": 0.04933315491324323,
      "grad_norm": 255.7532501220703,
      "learning_rate": 3e-06,
      "loss": -42.4297,
      "reward": 2.4352500438690186,
      "reward_std": 0.32932066917419434,
      "rewards/correctness_reward_func": 1.8333333134651184,
      "rewards/int_reward_func": 0.4791666716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.12275000661611557,
      "step": 553,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.0494223649582943,
      "grad_norm": 263.04913330078125,
      "learning_rate": 3e-06,
      "loss": -27.7121,
      "step": 554
    },
    {
      "epoch": 0.049511575003345375,
      "grad_norm": 187.12600708007812,
      "learning_rate": 3e-06,
      "loss": -32.293,
      "step": 555
    },
    {
      "epoch": 0.04960078504839645,
      "grad_norm": 233.73802185058594,
      "learning_rate": 3e-06,
      "loss": -34.2649,
      "step": 556
    },
    {
      "epoch": 0.04968999509344752,
      "grad_norm": 238.03567504882812,
      "learning_rate": 3e-06,
      "loss": -48.0373,
      "step": 557
    },
    {
      "epoch": 0.049779205138498595,
      "grad_norm": 211.17440795898438,
      "learning_rate": 3e-06,
      "loss": -34.5469,
      "step": 558
    },
    {
      "epoch": 0.04986841518354967,
      "grad_norm": 209.6473388671875,
      "learning_rate": 3e-06,
      "loss": -45.0123,
      "step": 559
    },
    {
      "epoch": 0.04995762522860074,
      "grad_norm": 219.1716766357422,
      "learning_rate": 3e-06,
      "loss": -29.3907,
      "step": 560
    },
    {
      "epoch": 0.050046835273651814,
      "grad_norm": 194.5946502685547,
      "learning_rate": 3e-06,
      "loss": -33.3813,
      "step": 561
    },
    {
      "epoch": 0.05013604531870289,
      "grad_norm": 230.82928466796875,
      "learning_rate": 3e-06,
      "loss": -37.409,
      "step": 562
    },
    {
      "epoch": 0.05022525536375396,
      "grad_norm": 268.3168640136719,
      "learning_rate": 3e-06,
      "loss": -50.4616,
      "step": 563
    },
    {
      "epoch": 0.050314465408805034,
      "grad_norm": 225.2816925048828,
      "learning_rate": 3e-06,
      "loss": -36.9552,
      "step": 564
    },
    {
      "completion_length": 138.81250381469727,
      "epoch": 0.05040367545385611,
      "grad_norm": 236.8852996826172,
      "learning_rate": 3e-06,
      "loss": 45.2003,
      "reward": 1.8692501783370972,
      "reward_std": 0.7652427852153778,
      "rewards/correctness_reward_func": 1.25,
      "rewards/int_reward_func": 0.46875,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.15049999579787254,
      "step": 565,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.050492885498907174,
      "grad_norm": 258.247802734375,
      "learning_rate": 3e-06,
      "loss": 49.6383,
      "step": 566
    },
    {
      "epoch": 0.05058209554395825,
      "grad_norm": 318.1617126464844,
      "learning_rate": 3e-06,
      "loss": 58.9489,
      "step": 567
    },
    {
      "epoch": 0.05067130558900932,
      "grad_norm": 226.23045349121094,
      "learning_rate": 3e-06,
      "loss": 44.3005,
      "step": 568
    },
    {
      "epoch": 0.050760515634060394,
      "grad_norm": 303.47760009765625,
      "learning_rate": 3e-06,
      "loss": 28.8653,
      "step": 569
    },
    {
      "epoch": 0.05084972567911147,
      "grad_norm": 248.53013610839844,
      "learning_rate": 3e-06,
      "loss": 45.7073,
      "step": 570
    },
    {
      "epoch": 0.05093893572416254,
      "grad_norm": 228.76365661621094,
      "learning_rate": 3e-06,
      "loss": 40.9719,
      "step": 571
    },
    {
      "epoch": 0.05102814576921361,
      "grad_norm": 236.98915100097656,
      "learning_rate": 3e-06,
      "loss": 44.3296,
      "step": 572
    },
    {
      "epoch": 0.05111735581426469,
      "grad_norm": 318.7423400878906,
      "learning_rate": 3e-06,
      "loss": 54.9659,
      "step": 573
    },
    {
      "epoch": 0.05120656585931576,
      "grad_norm": 226.2831268310547,
      "learning_rate": 3e-06,
      "loss": 40.0793,
      "step": 574
    },
    {
      "epoch": 0.05129577590436683,
      "grad_norm": 321.7300109863281,
      "learning_rate": 3e-06,
      "loss": 25.4049,
      "step": 575
    },
    {
      "epoch": 0.051384985949417906,
      "grad_norm": 225.1118927001953,
      "learning_rate": 3e-06,
      "loss": 42.5627,
      "step": 576
    },
    {
      "completion_length": 114.56250381469727,
      "epoch": 0.05147419599446898,
      "grad_norm": 151.06048583984375,
      "learning_rate": 3e-06,
      "loss": 33.113,
      "reward": 2.1812918186187744,
      "reward_std": 0.530484139919281,
      "rewards/correctness_reward_func": 1.5833333134651184,
      "rewards/int_reward_func": 0.4375,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.16045832633972168,
      "step": 577,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.05156340603952005,
      "grad_norm": 182.15431213378906,
      "learning_rate": 3e-06,
      "loss": 19.2473,
      "step": 578
    },
    {
      "epoch": 0.05165261608457112,
      "grad_norm": 169.3698272705078,
      "learning_rate": 3e-06,
      "loss": 25.8161,
      "step": 579
    },
    {
      "epoch": 0.05174182612962219,
      "grad_norm": 155.8734588623047,
      "learning_rate": 3e-06,
      "loss": 19.9518,
      "step": 580
    },
    {
      "epoch": 0.051831036174673266,
      "grad_norm": 177.54641723632812,
      "learning_rate": 3e-06,
      "loss": 23.4249,
      "step": 581
    },
    {
      "epoch": 0.05192024621972434,
      "grad_norm": 143.13719177246094,
      "learning_rate": 3e-06,
      "loss": 7.9228,
      "step": 582
    },
    {
      "epoch": 0.05200945626477541,
      "grad_norm": 146.39906311035156,
      "learning_rate": 3e-06,
      "loss": 30.8804,
      "step": 583
    },
    {
      "epoch": 0.052098666309826486,
      "grad_norm": 166.2614288330078,
      "learning_rate": 3e-06,
      "loss": 17.3062,
      "step": 584
    },
    {
      "epoch": 0.05218787635487756,
      "grad_norm": 158.1491241455078,
      "learning_rate": 3e-06,
      "loss": 23.1792,
      "step": 585
    },
    {
      "epoch": 0.05227708639992863,
      "grad_norm": 124.15723419189453,
      "learning_rate": 3e-06,
      "loss": 18.8908,
      "step": 586
    },
    {
      "epoch": 0.052366296444979706,
      "grad_norm": 155.26602172851562,
      "learning_rate": 3e-06,
      "loss": 21.968,
      "step": 587
    },
    {
      "epoch": 0.05245550649003078,
      "grad_norm": 155.0635528564453,
      "learning_rate": 3e-06,
      "loss": 6.3548,
      "step": 588
    },
    {
      "completion_length": 161.45833587646484,
      "epoch": 0.05254471653508185,
      "grad_norm": 317.9678649902344,
      "learning_rate": 3e-06,
      "loss": 6.4115,
      "reward": 1.8582292199134827,
      "reward_std": 0.48372724652290344,
      "rewards/correctness_reward_func": 1.3333333134651184,
      "rewards/int_reward_func": 0.4895833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.03531250241212547,
      "step": 589,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.052633926580132925,
      "grad_norm": 267.894287109375,
      "learning_rate": 3e-06,
      "loss": 6.0494,
      "step": 590
    },
    {
      "epoch": 0.052723136625184,
      "grad_norm": 197.32470703125,
      "learning_rate": 3e-06,
      "loss": 8.4095,
      "step": 591
    },
    {
      "epoch": 0.05281234667023507,
      "grad_norm": 208.84291076660156,
      "learning_rate": 3e-06,
      "loss": 2.183,
      "step": 592
    },
    {
      "epoch": 0.05290155671528614,
      "grad_norm": 213.50672912597656,
      "learning_rate": 3e-06,
      "loss": -5.6327,
      "step": 593
    },
    {
      "epoch": 0.05299076676033721,
      "grad_norm": 264.34210205078125,
      "learning_rate": 3e-06,
      "loss": 0.0463,
      "step": 594
    },
    {
      "epoch": 0.053079976805388285,
      "grad_norm": 269.38372802734375,
      "learning_rate": 3e-06,
      "loss": 3.4921,
      "step": 595
    },
    {
      "epoch": 0.05316918685043936,
      "grad_norm": 233.82005310058594,
      "learning_rate": 3e-06,
      "loss": 3.9645,
      "step": 596
    },
    {
      "epoch": 0.05325839689549043,
      "grad_norm": 174.2704620361328,
      "learning_rate": 3e-06,
      "loss": 6.6277,
      "step": 597
    },
    {
      "epoch": 0.053347606940541505,
      "grad_norm": 197.27203369140625,
      "learning_rate": 3e-06,
      "loss": -0.034,
      "step": 598
    },
    {
      "epoch": 0.05343681698559258,
      "grad_norm": 195.1741943359375,
      "learning_rate": 3e-06,
      "loss": -7.7007,
      "step": 599
    },
    {
      "epoch": 0.05352602703064365,
      "grad_norm": 218.35403442382812,
      "learning_rate": 3e-06,
      "loss": -2.732,
      "step": 600
    },
    {
      "completion_length": 163.20833587646484,
      "epoch": 0.053615237075694724,
      "grad_norm": 89.40003204345703,
      "learning_rate": 3e-06,
      "loss": -15.5412,
      "reward": 1.958250105381012,
      "reward_std": 0.20268601924180984,
      "rewards/correctness_reward_func": 1.4583333134651184,
      "rewards/int_reward_func": 0.4375,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.06241666525602341,
      "step": 601,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.0537044471207458,
      "grad_norm": 85.92318725585938,
      "learning_rate": 3e-06,
      "loss": -19.9741,
      "step": 602
    },
    {
      "epoch": 0.05379365716579687,
      "grad_norm": 89.12326049804688,
      "learning_rate": 3e-06,
      "loss": -28.9968,
      "step": 603
    },
    {
      "epoch": 0.053882867210847944,
      "grad_norm": 109.64755249023438,
      "learning_rate": 3e-06,
      "loss": -21.6802,
      "step": 604
    },
    {
      "epoch": 0.05397207725589902,
      "grad_norm": 99.89476776123047,
      "learning_rate": 3e-06,
      "loss": -16.4453,
      "step": 605
    },
    {
      "epoch": 0.054061287300950084,
      "grad_norm": 140.71066284179688,
      "learning_rate": 3e-06,
      "loss": -24.1026,
      "step": 606
    },
    {
      "epoch": 0.05415049734600116,
      "grad_norm": 80.75763702392578,
      "learning_rate": 3e-06,
      "loss": -16.613,
      "step": 607
    },
    {
      "epoch": 0.05423970739105223,
      "grad_norm": 85.42610168457031,
      "learning_rate": 3e-06,
      "loss": -21.2449,
      "step": 608
    },
    {
      "epoch": 0.054328917436103304,
      "grad_norm": 93.39994812011719,
      "learning_rate": 3e-06,
      "loss": -30.1845,
      "step": 609
    },
    {
      "epoch": 0.05441812748115438,
      "grad_norm": 96.1513671875,
      "learning_rate": 3e-06,
      "loss": -22.8781,
      "step": 610
    },
    {
      "epoch": 0.05450733752620545,
      "grad_norm": 98.65193176269531,
      "learning_rate": 3e-06,
      "loss": -17.6772,
      "step": 611
    },
    {
      "epoch": 0.05459654757125652,
      "grad_norm": 130.4186248779297,
      "learning_rate": 3e-06,
      "loss": -25.9751,
      "step": 612
    },
    {
      "completion_length": 129.4375,
      "epoch": 0.0546857576163076,
      "grad_norm": 178.97259521484375,
      "learning_rate": 3e-06,
      "loss": -41.7741,
      "reward": 1.9304792881011963,
      "reward_std": 0.4192664921283722,
      "rewards/correctness_reward_func": 1.2916666865348816,
      "rewards/int_reward_func": 0.4895833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.14922916889190674,
      "step": 613,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.05477496766135867,
      "grad_norm": 159.00680541992188,
      "learning_rate": 3e-06,
      "loss": -36.4024,
      "step": 614
    },
    {
      "epoch": 0.05486417770640974,
      "grad_norm": 154.65304565429688,
      "learning_rate": 3e-06,
      "loss": -35.2622,
      "step": 615
    },
    {
      "epoch": 0.05495338775146082,
      "grad_norm": 239.50408935546875,
      "learning_rate": 3e-06,
      "loss": -34.9883,
      "step": 616
    },
    {
      "epoch": 0.05504259779651189,
      "grad_norm": 191.45263671875,
      "learning_rate": 3e-06,
      "loss": -30.4894,
      "step": 617
    },
    {
      "epoch": 0.05513180784156296,
      "grad_norm": 172.67025756835938,
      "learning_rate": 3e-06,
      "loss": -40.9277,
      "step": 618
    },
    {
      "epoch": 0.05522101788661403,
      "grad_norm": 180.4842071533203,
      "learning_rate": 3e-06,
      "loss": -44.2844,
      "step": 619
    },
    {
      "epoch": 0.0553102279316651,
      "grad_norm": 175.10528564453125,
      "learning_rate": 3e-06,
      "loss": -39.7696,
      "step": 620
    },
    {
      "epoch": 0.055399437976716176,
      "grad_norm": 224.33847045898438,
      "learning_rate": 3e-06,
      "loss": -38.3443,
      "step": 621
    },
    {
      "epoch": 0.05548864802176725,
      "grad_norm": 196.98231506347656,
      "learning_rate": 3e-06,
      "loss": -36.4789,
      "step": 622
    },
    {
      "epoch": 0.05557785806681832,
      "grad_norm": 205.22146606445312,
      "learning_rate": 3e-06,
      "loss": -32.5847,
      "step": 623
    },
    {
      "epoch": 0.055667068111869396,
      "grad_norm": 189.9784393310547,
      "learning_rate": 3e-06,
      "loss": -43.7861,
      "step": 624
    },
    {
      "completion_length": 135.1666717529297,
      "epoch": 0.05575627815692047,
      "grad_norm": 190.61593627929688,
      "learning_rate": 3e-06,
      "loss": -23.515,
      "reward": 1.7549793124198914,
      "reward_std": 0.5575149804353714,
      "rewards/correctness_reward_func": 1.1666666865348816,
      "rewards/int_reward_func": 0.4479166716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.14039582759141922,
      "step": 625,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.05584548820197154,
      "grad_norm": 263.34173583984375,
      "learning_rate": 3e-06,
      "loss": -26.4535,
      "step": 626
    },
    {
      "epoch": 0.055934698247022616,
      "grad_norm": 203.12969970703125,
      "learning_rate": 3e-06,
      "loss": -20.2943,
      "step": 627
    },
    {
      "epoch": 0.05602390829207369,
      "grad_norm": 186.3466033935547,
      "learning_rate": 3e-06,
      "loss": -22.5875,
      "step": 628
    },
    {
      "epoch": 0.05611311833712476,
      "grad_norm": 206.43478393554688,
      "learning_rate": 3e-06,
      "loss": -14.8606,
      "step": 629
    },
    {
      "epoch": 0.056202328382175835,
      "grad_norm": 230.95394897460938,
      "learning_rate": 3e-06,
      "loss": -11.2027,
      "step": 630
    },
    {
      "epoch": 0.05629153842722691,
      "grad_norm": 208.40184020996094,
      "learning_rate": 3e-06,
      "loss": -24.9903,
      "step": 631
    },
    {
      "epoch": 0.056380748472277975,
      "grad_norm": 283.0361328125,
      "learning_rate": 3e-06,
      "loss": -28.8647,
      "step": 632
    },
    {
      "epoch": 0.05646995851732905,
      "grad_norm": 243.43634033203125,
      "learning_rate": 3e-06,
      "loss": -20.9854,
      "step": 633
    },
    {
      "epoch": 0.05655916856238012,
      "grad_norm": 196.8306121826172,
      "learning_rate": 3e-06,
      "loss": -25.1397,
      "step": 634
    },
    {
      "epoch": 0.056648378607431195,
      "grad_norm": 204.37130737304688,
      "learning_rate": 3e-06,
      "loss": -17.3989,
      "step": 635
    },
    {
      "epoch": 0.05673758865248227,
      "grad_norm": 222.9701385498047,
      "learning_rate": 3e-06,
      "loss": -14.9143,
      "step": 636
    },
    {
      "completion_length": 117.18750381469727,
      "epoch": 0.05682679869753334,
      "grad_norm": 152.56382751464844,
      "learning_rate": 3e-06,
      "loss": -53.4148,
      "reward": 2.4275625944137573,
      "reward_std": 0.3748088702559471,
      "rewards/correctness_reward_func": 1.75,
      "rewards/int_reward_func": 0.4895833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.18797916173934937,
      "step": 637,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.056916008742584415,
      "grad_norm": 187.37356567382812,
      "learning_rate": 3e-06,
      "loss": -49.109,
      "step": 638
    },
    {
      "epoch": 0.05700521878763549,
      "grad_norm": 188.10548400878906,
      "learning_rate": 3e-06,
      "loss": -49.7143,
      "step": 639
    },
    {
      "epoch": 0.05709442883268656,
      "grad_norm": 152.7540283203125,
      "learning_rate": 3e-06,
      "loss": -46.4195,
      "step": 640
    },
    {
      "epoch": 0.057183638877737634,
      "grad_norm": 191.8466796875,
      "learning_rate": 3e-06,
      "loss": -54.4961,
      "step": 641
    },
    {
      "epoch": 0.05727284892278871,
      "grad_norm": 160.72772216796875,
      "learning_rate": 3e-06,
      "loss": -46.0513,
      "step": 642
    },
    {
      "epoch": 0.05736205896783978,
      "grad_norm": 163.2805938720703,
      "learning_rate": 3e-06,
      "loss": -55.7387,
      "step": 643
    },
    {
      "epoch": 0.057451269012890854,
      "grad_norm": 189.55470275878906,
      "learning_rate": 3e-06,
      "loss": -51.4447,
      "step": 644
    },
    {
      "epoch": 0.05754047905794193,
      "grad_norm": 182.82583618164062,
      "learning_rate": 3e-06,
      "loss": -51.1561,
      "step": 645
    },
    {
      "epoch": 0.057629689102992994,
      "grad_norm": 177.5033721923828,
      "learning_rate": 3e-06,
      "loss": -49.5066,
      "step": 646
    },
    {
      "epoch": 0.05771889914804407,
      "grad_norm": 208.1852569580078,
      "learning_rate": 3e-06,
      "loss": -59.1063,
      "step": 647
    },
    {
      "epoch": 0.05780810919309514,
      "grad_norm": 185.0785369873047,
      "learning_rate": 3e-06,
      "loss": -49.7641,
      "step": 648
    },
    {
      "completion_length": 127.75000762939453,
      "epoch": 0.057897319238146214,
      "grad_norm": 332.4126892089844,
      "learning_rate": 3e-06,
      "loss": 24.9465,
      "reward": 2.10916668176651,
      "reward_std": 0.7774414718151093,
      "rewards/correctness_reward_func": 1.5,
      "rewards/int_reward_func": 0.4583333283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.15083333104848862,
      "step": 649,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.05798652928319729,
      "grad_norm": 274.673828125,
      "learning_rate": 3e-06,
      "loss": 19.6904,
      "step": 650
    },
    {
      "epoch": 0.05807573932824836,
      "grad_norm": 225.66183471679688,
      "learning_rate": 3e-06,
      "loss": 17.079,
      "step": 651
    },
    {
      "epoch": 0.058164949373299434,
      "grad_norm": 236.45230102539062,
      "learning_rate": 3e-06,
      "loss": 16.4363,
      "step": 652
    },
    {
      "epoch": 0.05825415941835051,
      "grad_norm": 304.5491943359375,
      "learning_rate": 3e-06,
      "loss": 33.7894,
      "step": 653
    },
    {
      "epoch": 0.05834336946340158,
      "grad_norm": 294.2102966308594,
      "learning_rate": 3e-06,
      "loss": 12.1637,
      "step": 654
    },
    {
      "epoch": 0.05843257950845265,
      "grad_norm": 307.0853271484375,
      "learning_rate": 3e-06,
      "loss": 23.9968,
      "step": 655
    },
    {
      "epoch": 0.05852178955350373,
      "grad_norm": 267.52960205078125,
      "learning_rate": 3e-06,
      "loss": 18.2069,
      "step": 656
    },
    {
      "epoch": 0.0586109995985548,
      "grad_norm": 219.6736602783203,
      "learning_rate": 3e-06,
      "loss": 16.1471,
      "step": 657
    },
    {
      "epoch": 0.05870020964360587,
      "grad_norm": 312.0810852050781,
      "learning_rate": 3e-06,
      "loss": 15.3629,
      "step": 658
    },
    {
      "epoch": 0.05878941968865694,
      "grad_norm": 355.7622985839844,
      "learning_rate": 3e-06,
      "loss": 32.1285,
      "step": 659
    },
    {
      "epoch": 0.05887862973370801,
      "grad_norm": 243.60047912597656,
      "learning_rate": 3e-06,
      "loss": 10.9719,
      "step": 660
    },
    {
      "completion_length": 128.37500381469727,
      "epoch": 0.058967839778759086,
      "grad_norm": 160.17491149902344,
      "learning_rate": 3e-06,
      "loss": -4.7968,
      "reward": 2.220729112625122,
      "reward_std": 0.37828393280506134,
      "rewards/correctness_reward_func": 1.5833333134651184,
      "rewards/int_reward_func": 0.5,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.13739583641290665,
      "step": 661,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.05905704982381016,
      "grad_norm": 213.04843139648438,
      "learning_rate": 3e-06,
      "loss": -17.5158,
      "step": 662
    },
    {
      "epoch": 0.05914625986886123,
      "grad_norm": 360.1976318359375,
      "learning_rate": 3e-06,
      "loss": -5.8696,
      "step": 663
    },
    {
      "epoch": 0.059235469913912306,
      "grad_norm": 165.9031524658203,
      "learning_rate": 3e-06,
      "loss": 9.1976,
      "step": 664
    },
    {
      "epoch": 0.05932467995896338,
      "grad_norm": 185.02491760253906,
      "learning_rate": 3e-06,
      "loss": 0.8362,
      "step": 665
    },
    {
      "epoch": 0.05941389000401445,
      "grad_norm": 186.4868927001953,
      "learning_rate": 3e-06,
      "loss": -4.8959,
      "step": 666
    },
    {
      "epoch": 0.059503100049065526,
      "grad_norm": 172.38906860351562,
      "learning_rate": 3e-06,
      "loss": -8.3078,
      "step": 667
    },
    {
      "epoch": 0.0595923100941166,
      "grad_norm": 205.17637634277344,
      "learning_rate": 3e-06,
      "loss": -19.4471,
      "step": 668
    },
    {
      "epoch": 0.05968152013916767,
      "grad_norm": 412.4108581542969,
      "learning_rate": 3e-06,
      "loss": -9.0223,
      "step": 669
    },
    {
      "epoch": 0.059770730184218746,
      "grad_norm": 165.48020935058594,
      "learning_rate": 3e-06,
      "loss": 5.0711,
      "step": 670
    },
    {
      "epoch": 0.05985994022926982,
      "grad_norm": 185.8058624267578,
      "learning_rate": 3e-06,
      "loss": -3.3781,
      "step": 671
    },
    {
      "epoch": 0.059949150274320885,
      "grad_norm": 214.62686157226562,
      "learning_rate": 3e-06,
      "loss": -9.8208,
      "step": 672
    },
    {
      "completion_length": 110.20833587646484,
      "epoch": 0.06003836031937196,
      "grad_norm": 223.97824096679688,
      "learning_rate": 3e-06,
      "loss": 17.6495,
      "reward": 1.9757083654403687,
      "reward_std": 0.7511122822761536,
      "rewards/correctness_reward_func": 1.375,
      "rewards/int_reward_func": 0.4375,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.16320832818746567,
      "step": 673,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.06012757036442303,
      "grad_norm": 180.975341796875,
      "learning_rate": 3e-06,
      "loss": -3.9399,
      "step": 674
    },
    {
      "epoch": 0.060216780409474105,
      "grad_norm": 222.6065216064453,
      "learning_rate": 3e-06,
      "loss": 10.3809,
      "step": 675
    },
    {
      "epoch": 0.06030599045452518,
      "grad_norm": 192.075439453125,
      "learning_rate": 3e-06,
      "loss": 14.9998,
      "step": 676
    },
    {
      "epoch": 0.06039520049957625,
      "grad_norm": 200.7101593017578,
      "learning_rate": 3e-06,
      "loss": 2.2906,
      "step": 677
    },
    {
      "epoch": 0.060484410544627325,
      "grad_norm": 200.1593017578125,
      "learning_rate": 3e-06,
      "loss": 13.137,
      "step": 678
    },
    {
      "epoch": 0.0605736205896784,
      "grad_norm": 185.83168029785156,
      "learning_rate": 3e-06,
      "loss": 16.2497,
      "step": 679
    },
    {
      "epoch": 0.06066283063472947,
      "grad_norm": 208.6911163330078,
      "learning_rate": 3e-06,
      "loss": -5.308,
      "step": 680
    },
    {
      "epoch": 0.060752040679780545,
      "grad_norm": 231.81312561035156,
      "learning_rate": 3e-06,
      "loss": 8.1785,
      "step": 681
    },
    {
      "epoch": 0.06084125072483162,
      "grad_norm": 187.45535278320312,
      "learning_rate": 3e-06,
      "loss": 12.5542,
      "step": 682
    },
    {
      "epoch": 0.06093046076988269,
      "grad_norm": 182.04257202148438,
      "learning_rate": 3e-06,
      "loss": -1.0645,
      "step": 683
    },
    {
      "epoch": 0.061019670814933764,
      "grad_norm": 186.71937561035156,
      "learning_rate": 3e-06,
      "loss": 10.3416,
      "step": 684
    },
    {
      "completion_length": 130.4791717529297,
      "epoch": 0.06110888085998484,
      "grad_norm": 169.67178344726562,
      "learning_rate": 3e-06,
      "loss": 32.9847,
      "reward": 2.2736042737960815,
      "reward_std": 0.5084125399589539,
      "rewards/correctness_reward_func": 1.6666666865348816,
      "rewards/int_reward_func": 0.4583333283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.14860416948795319,
      "step": 685,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.061198090905035904,
      "grad_norm": 231.2995147705078,
      "learning_rate": 3e-06,
      "loss": 18.3002,
      "step": 686
    },
    {
      "epoch": 0.06128730095008698,
      "grad_norm": 162.75083923339844,
      "learning_rate": 3e-06,
      "loss": 29.4583,
      "step": 687
    },
    {
      "epoch": 0.06137651099513805,
      "grad_norm": 209.54966735839844,
      "learning_rate": 3e-06,
      "loss": 42.4661,
      "step": 688
    },
    {
      "epoch": 0.061465721040189124,
      "grad_norm": 396.1800537109375,
      "learning_rate": 3e-06,
      "loss": 42.6963,
      "step": 689
    },
    {
      "epoch": 0.0615549310852402,
      "grad_norm": 144.02049255371094,
      "learning_rate": 3e-06,
      "loss": 23.2087,
      "step": 690
    },
    {
      "epoch": 0.06164414113029127,
      "grad_norm": 179.72213745117188,
      "learning_rate": 3e-06,
      "loss": 28.8186,
      "step": 691
    },
    {
      "epoch": 0.061733351175342344,
      "grad_norm": 143.9263153076172,
      "learning_rate": 3e-06,
      "loss": 16.2638,
      "step": 692
    },
    {
      "epoch": 0.06182256122039342,
      "grad_norm": 148.50750732421875,
      "learning_rate": 3e-06,
      "loss": 27.5098,
      "step": 693
    },
    {
      "epoch": 0.06191177126544449,
      "grad_norm": 178.09515380859375,
      "learning_rate": 3e-06,
      "loss": 38.1567,
      "step": 694
    },
    {
      "epoch": 0.06200098131049556,
      "grad_norm": 319.7694396972656,
      "learning_rate": 3e-06,
      "loss": 36.9073,
      "step": 695
    },
    {
      "epoch": 0.06209019135554664,
      "grad_norm": 137.5644989013672,
      "learning_rate": 3e-06,
      "loss": 19.9691,
      "step": 696
    },
    {
      "completion_length": 138.20833587646484,
      "epoch": 0.06217940140059771,
      "grad_norm": 243.18804931640625,
      "learning_rate": 3e-06,
      "loss": 63.9782,
      "reward": 1.6311666369438171,
      "reward_std": 0.6314830482006073,
      "rewards/correctness_reward_func": 1.0833333730697632,
      "rewards/int_reward_func": 0.4166666716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.13116667047142982,
      "step": 697,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.06226861144564878,
      "grad_norm": 248.19979858398438,
      "learning_rate": 3e-06,
      "loss": 62.0708,
      "step": 698
    },
    {
      "epoch": 0.06235782149069985,
      "grad_norm": 192.6903839111328,
      "learning_rate": 3e-06,
      "loss": 44.2788,
      "step": 699
    },
    {
      "epoch": 0.06244703153575092,
      "grad_norm": 187.61729431152344,
      "learning_rate": 3e-06,
      "loss": 45.635,
      "step": 700
    },
    {
      "epoch": 0.062536241580802,
      "grad_norm": 270.56439208984375,
      "learning_rate": 3e-06,
      "loss": 61.8371,
      "step": 701
    },
    {
      "epoch": 0.06262545162585308,
      "grad_norm": 186.34654235839844,
      "learning_rate": 3e-06,
      "loss": 35.0754,
      "step": 702
    },
    {
      "epoch": 0.06271466167090414,
      "grad_norm": 193.72666931152344,
      "learning_rate": 3e-06,
      "loss": 56.4102,
      "step": 703
    },
    {
      "epoch": 0.06280387171595522,
      "grad_norm": 213.4738006591797,
      "learning_rate": 3e-06,
      "loss": 55.6052,
      "step": 704
    },
    {
      "epoch": 0.06289308176100629,
      "grad_norm": 178.9663543701172,
      "learning_rate": 3e-06,
      "loss": 38.9399,
      "step": 705
    },
    {
      "epoch": 0.06298229180605736,
      "grad_norm": 155.2235870361328,
      "learning_rate": 3e-06,
      "loss": 40.2563,
      "step": 706
    },
    {
      "epoch": 0.06307150185110844,
      "grad_norm": 190.83424377441406,
      "learning_rate": 3e-06,
      "loss": 54.1227,
      "step": 707
    },
    {
      "epoch": 0.0631607118961595,
      "grad_norm": 151.27175903320312,
      "learning_rate": 3e-06,
      "loss": 29.2768,
      "step": 708
    },
    {
      "completion_length": 177.89584350585938,
      "epoch": 0.06324992194121058,
      "grad_norm": 383.90155029296875,
      "learning_rate": 3e-06,
      "loss": -57.4567,
      "reward": 1.6143542528152466,
      "reward_std": 1.1507561206817627,
      "rewards/correctness_reward_func": 1.125,
      "rewards/int_reward_func": 0.4479166716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.041437502950429916,
      "step": 709,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.06333913198626165,
      "grad_norm": 228.4014892578125,
      "learning_rate": 3e-06,
      "loss": -22.536,
      "step": 710
    },
    {
      "epoch": 0.06342834203131273,
      "grad_norm": 247.7998809814453,
      "learning_rate": 3e-06,
      "loss": -34.7459,
      "step": 711
    },
    {
      "epoch": 0.0635175520763638,
      "grad_norm": 256.2286376953125,
      "learning_rate": 3e-06,
      "loss": -30.4245,
      "step": 712
    },
    {
      "epoch": 0.06360676212141488,
      "grad_norm": 254.9169158935547,
      "learning_rate": 3e-06,
      "loss": -38.2843,
      "step": 713
    },
    {
      "epoch": 0.06369597216646594,
      "grad_norm": 321.43609619140625,
      "learning_rate": 3e-06,
      "loss": -35.5213,
      "step": 714
    },
    {
      "epoch": 0.06378518221151702,
      "grad_norm": 349.0517272949219,
      "learning_rate": 3e-06,
      "loss": -55.9526,
      "step": 715
    },
    {
      "epoch": 0.06387439225656809,
      "grad_norm": 209.25282287597656,
      "learning_rate": 3e-06,
      "loss": -22.3323,
      "step": 716
    },
    {
      "epoch": 0.06396360230161917,
      "grad_norm": 247.7156219482422,
      "learning_rate": 3e-06,
      "loss": -35.9537,
      "step": 717
    },
    {
      "epoch": 0.06405281234667023,
      "grad_norm": 274.3576965332031,
      "learning_rate": 3e-06,
      "loss": -33.4144,
      "step": 718
    },
    {
      "epoch": 0.0641420223917213,
      "grad_norm": 287.4893798828125,
      "learning_rate": 3e-06,
      "loss": -40.4848,
      "step": 719
    },
    {
      "epoch": 0.06423123243677238,
      "grad_norm": 375.7614440917969,
      "learning_rate": 3e-06,
      "loss": -39.3635,
      "step": 720
    },
    {
      "completion_length": 109.22916793823242,
      "epoch": 0.06432044248182345,
      "grad_norm": 123.53705596923828,
      "learning_rate": 3e-06,
      "loss": -30.919,
      "reward": 2.3711042404174805,
      "reward_std": 0.4830681085586548,
      "rewards/correctness_reward_func": 1.7083333730697632,
      "rewards/int_reward_func": 0.46875,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.1940208300948143,
      "step": 721,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.06440965252687453,
      "grad_norm": 114.69866943359375,
      "learning_rate": 3e-06,
      "loss": -29.2645,
      "step": 722
    },
    {
      "epoch": 0.0644988625719256,
      "grad_norm": 118.97370910644531,
      "learning_rate": 3e-06,
      "loss": -25.0985,
      "step": 723
    },
    {
      "epoch": 0.06458807261697667,
      "grad_norm": 121.6566162109375,
      "learning_rate": 3e-06,
      "loss": -29.0591,
      "step": 724
    },
    {
      "epoch": 0.06467728266202774,
      "grad_norm": 182.92691040039062,
      "learning_rate": 3e-06,
      "loss": -28.0901,
      "step": 725
    },
    {
      "epoch": 0.06476649270707882,
      "grad_norm": 156.50718688964844,
      "learning_rate": 3e-06,
      "loss": -30.7831,
      "step": 726
    },
    {
      "epoch": 0.06485570275212989,
      "grad_norm": 132.53089904785156,
      "learning_rate": 3e-06,
      "loss": -33.3058,
      "step": 727
    },
    {
      "epoch": 0.06494491279718097,
      "grad_norm": 112.21791076660156,
      "learning_rate": 3e-06,
      "loss": -31.8015,
      "step": 728
    },
    {
      "epoch": 0.06503412284223203,
      "grad_norm": 130.97052001953125,
      "learning_rate": 3e-06,
      "loss": -27.7953,
      "step": 729
    },
    {
      "epoch": 0.06512333288728311,
      "grad_norm": 128.9853515625,
      "learning_rate": 3e-06,
      "loss": -32.2601,
      "step": 730
    },
    {
      "epoch": 0.06521254293233418,
      "grad_norm": 146.0636444091797,
      "learning_rate": 3e-06,
      "loss": -31.9068,
      "step": 731
    },
    {
      "epoch": 0.06530175297738525,
      "grad_norm": 148.94358825683594,
      "learning_rate": 3e-06,
      "loss": -35.98,
      "step": 732
    },
    {
      "completion_length": 133.77083587646484,
      "epoch": 0.06539096302243633,
      "grad_norm": 200.87583923339844,
      "learning_rate": 3e-06,
      "loss": 5.8156,
      "reward": 1.7448542714118958,
      "reward_std": 0.7897588908672333,
      "rewards/correctness_reward_func": 1.1666666865348816,
      "rewards/int_reward_func": 0.4479166716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.1302708424627781,
      "step": 733,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.0654801730674874,
      "grad_norm": 254.33050537109375,
      "learning_rate": 3e-06,
      "loss": 39.6914,
      "step": 734
    },
    {
      "epoch": 0.06556938311253847,
      "grad_norm": 207.40354919433594,
      "learning_rate": 3e-06,
      "loss": 18.8055,
      "step": 735
    },
    {
      "epoch": 0.06565859315758954,
      "grad_norm": 255.23114013671875,
      "learning_rate": 3e-06,
      "loss": 12.5947,
      "step": 736
    },
    {
      "epoch": 0.06574780320264062,
      "grad_norm": 183.82200622558594,
      "learning_rate": 3e-06,
      "loss": 11.0044,
      "step": 737
    },
    {
      "epoch": 0.06583701324769169,
      "grad_norm": 226.2420654296875,
      "learning_rate": 3e-06,
      "loss": 0.4847,
      "step": 738
    },
    {
      "epoch": 0.06592622329274277,
      "grad_norm": 233.6065673828125,
      "learning_rate": 3e-06,
      "loss": 4.5426,
      "step": 739
    },
    {
      "epoch": 0.06601543333779383,
      "grad_norm": 264.205078125,
      "learning_rate": 3e-06,
      "loss": 39.3686,
      "step": 740
    },
    {
      "epoch": 0.06610464338284491,
      "grad_norm": 241.85284423828125,
      "learning_rate": 3e-06,
      "loss": 17.7268,
      "step": 741
    },
    {
      "epoch": 0.06619385342789598,
      "grad_norm": 221.0516357421875,
      "learning_rate": 3e-06,
      "loss": 13.2982,
      "step": 742
    },
    {
      "epoch": 0.06628306347294706,
      "grad_norm": 212.37222290039062,
      "learning_rate": 3e-06,
      "loss": 7.7467,
      "step": 743
    },
    {
      "epoch": 0.06637227351799813,
      "grad_norm": 263.42919921875,
      "learning_rate": 3e-06,
      "loss": -2.1754,
      "step": 744
    },
    {
      "completion_length": 125.64583587646484,
      "epoch": 0.0664614835630492,
      "grad_norm": 193.34835815429688,
      "learning_rate": 3e-06,
      "loss": 40.2546,
      "reward": 1.7169584035873413,
      "reward_std": 0.4941745698451996,
      "rewards/correctness_reward_func": 1.0833333730697632,
      "rewards/int_reward_func": 0.4791666567325592,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.15445833653211594,
      "step": 745,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.06655069360810027,
      "grad_norm": 212.19464111328125,
      "learning_rate": 3e-06,
      "loss": 41.4745,
      "step": 746
    },
    {
      "epoch": 0.06663990365315134,
      "grad_norm": 237.55845642089844,
      "learning_rate": 3e-06,
      "loss": 47.5379,
      "step": 747
    },
    {
      "epoch": 0.06672911369820242,
      "grad_norm": 185.4857635498047,
      "learning_rate": 3e-06,
      "loss": 43.0404,
      "step": 748
    },
    {
      "epoch": 0.06681832374325349,
      "grad_norm": 158.21678161621094,
      "learning_rate": 3e-06,
      "loss": 32.8558,
      "step": 749
    },
    {
      "epoch": 0.06690753378830457,
      "grad_norm": 212.29397583007812,
      "learning_rate": 3e-06,
      "loss": 41.9557,
      "step": 750
    },
    {
      "epoch": 0.06699674383335563,
      "grad_norm": 181.2362060546875,
      "learning_rate": 3e-06,
      "loss": 38.4185,
      "step": 751
    },
    {
      "epoch": 0.06708595387840671,
      "grad_norm": 186.73841857910156,
      "learning_rate": 3e-06,
      "loss": 37.1992,
      "step": 752
    },
    {
      "epoch": 0.06717516392345778,
      "grad_norm": 182.0499267578125,
      "learning_rate": 3e-06,
      "loss": 42.4944,
      "step": 753
    },
    {
      "epoch": 0.06726437396850886,
      "grad_norm": 161.4265899658203,
      "learning_rate": 3e-06,
      "loss": 40.1143,
      "step": 754
    },
    {
      "epoch": 0.06735358401355993,
      "grad_norm": 145.66175842285156,
      "learning_rate": 3e-06,
      "loss": 29.066,
      "step": 755
    },
    {
      "epoch": 0.067442794058611,
      "grad_norm": 188.43362426757812,
      "learning_rate": 3e-06,
      "loss": 34.8647,
      "step": 756
    },
    {
      "completion_length": 115.64583587646484,
      "epoch": 0.06753200410366207,
      "grad_norm": 176.6951446533203,
      "learning_rate": 3e-06,
      "loss": 23.7273,
      "reward": 1.9728541374206543,
      "reward_std": 0.22583025321364403,
      "rewards/correctness_reward_func": 1.375,
      "rewards/int_reward_func": 0.4270833432674408,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.17077083885669708,
      "step": 757,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.06762121414871315,
      "grad_norm": 134.009521484375,
      "learning_rate": 3e-06,
      "loss": 29.2333,
      "step": 758
    },
    {
      "epoch": 0.06771042419376422,
      "grad_norm": 131.3091278076172,
      "learning_rate": 3e-06,
      "loss": 29.9313,
      "step": 759
    },
    {
      "epoch": 0.06779963423881528,
      "grad_norm": 128.05767822265625,
      "learning_rate": 3e-06,
      "loss": 22.7093,
      "step": 760
    },
    {
      "epoch": 0.06788884428386636,
      "grad_norm": 115.9305419921875,
      "learning_rate": 3e-06,
      "loss": 23.3578,
      "step": 761
    },
    {
      "epoch": 0.06797805432891743,
      "grad_norm": 127.96971893310547,
      "learning_rate": 3e-06,
      "loss": 26.8069,
      "step": 762
    },
    {
      "epoch": 0.06806726437396851,
      "grad_norm": 172.38279724121094,
      "learning_rate": 3e-06,
      "loss": 20.0156,
      "step": 763
    },
    {
      "epoch": 0.06815647441901958,
      "grad_norm": 122.19217681884766,
      "learning_rate": 3e-06,
      "loss": 23.256,
      "step": 764
    },
    {
      "epoch": 0.06824568446407066,
      "grad_norm": 98.1166763305664,
      "learning_rate": 3e-06,
      "loss": 25.2087,
      "step": 765
    },
    {
      "epoch": 0.06833489450912172,
      "grad_norm": 104.1299819946289,
      "learning_rate": 3e-06,
      "loss": 18.9453,
      "step": 766
    },
    {
      "epoch": 0.0684241045541728,
      "grad_norm": 108.13124084472656,
      "learning_rate": 3e-06,
      "loss": 18.3841,
      "step": 767
    },
    {
      "epoch": 0.06851331459922387,
      "grad_norm": 107.33203887939453,
      "learning_rate": 3e-06,
      "loss": 21.8827,
      "step": 768
    },
    {
      "completion_length": 140.7291717529297,
      "epoch": 0.06860252464427495,
      "grad_norm": 176.08241271972656,
      "learning_rate": 3e-06,
      "loss": 1.9837,
      "reward": 2.0843957662582397,
      "reward_std": 0.670438677072525,
      "rewards/correctness_reward_func": 1.5,
      "rewards/int_reward_func": 0.46875,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.11564583331346512,
      "step": 769,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.06869173468932602,
      "grad_norm": 184.2529754638672,
      "learning_rate": 3e-06,
      "loss": -9.1612,
      "step": 770
    },
    {
      "epoch": 0.0687809447343771,
      "grad_norm": 124.57987213134766,
      "learning_rate": 3e-06,
      "loss": -1.167,
      "step": 771
    },
    {
      "epoch": 0.06887015477942816,
      "grad_norm": 146.37869262695312,
      "learning_rate": 3e-06,
      "loss": 7.9908,
      "step": 772
    },
    {
      "epoch": 0.06895936482447923,
      "grad_norm": 159.84788513183594,
      "learning_rate": 3e-06,
      "loss": -7.8619,
      "step": 773
    },
    {
      "epoch": 0.06904857486953031,
      "grad_norm": 165.3255157470703,
      "learning_rate": 3e-06,
      "loss": -9.9255,
      "step": 774
    },
    {
      "epoch": 0.06913778491458138,
      "grad_norm": 147.72352600097656,
      "learning_rate": 3e-06,
      "loss": 0.4622,
      "step": 775
    },
    {
      "epoch": 0.06922699495963246,
      "grad_norm": 161.2108917236328,
      "learning_rate": 3e-06,
      "loss": -9.3763,
      "step": 776
    },
    {
      "epoch": 0.06931620500468352,
      "grad_norm": 117.9613265991211,
      "learning_rate": 3e-06,
      "loss": -2.9584,
      "step": 777
    },
    {
      "epoch": 0.0694054150497346,
      "grad_norm": 127.0103988647461,
      "learning_rate": 3e-06,
      "loss": 4.5212,
      "step": 778
    },
    {
      "epoch": 0.06949462509478567,
      "grad_norm": 161.12701416015625,
      "learning_rate": 3e-06,
      "loss": -7.3502,
      "step": 779
    },
    {
      "epoch": 0.06958383513983675,
      "grad_norm": 147.03277587890625,
      "learning_rate": 3e-06,
      "loss": -12.0696,
      "step": 780
    },
    {
      "completion_length": 146.75000762939453,
      "epoch": 0.06967304518488782,
      "grad_norm": 93.77994537353516,
      "learning_rate": 3e-06,
      "loss": -9.5819,
      "reward": 2.0504584312438965,
      "reward_std": 0.46765226125717163,
      "rewards/correctness_reward_func": 1.4166666865348816,
      "rewards/int_reward_func": 0.5,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.010416666977107525,
      "rewards/xmlcount_reward_func": 0.12337498925626278,
      "step": 781,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.0697622552299389,
      "grad_norm": 109.34754180908203,
      "learning_rate": 3e-06,
      "loss": -14.1501,
      "step": 782
    },
    {
      "epoch": 0.06985146527498996,
      "grad_norm": 83.16600799560547,
      "learning_rate": 3e-06,
      "loss": -17.3924,
      "step": 783
    },
    {
      "epoch": 0.06994067532004104,
      "grad_norm": 83.6520767211914,
      "learning_rate": 3e-06,
      "loss": -3.3917,
      "step": 784
    },
    {
      "epoch": 0.07002988536509211,
      "grad_norm": 111.59048461914062,
      "learning_rate": 3e-06,
      "loss": -18.6001,
      "step": 785
    },
    {
      "epoch": 0.07011909541014318,
      "grad_norm": 81.26487731933594,
      "learning_rate": 3e-06,
      "loss": -4.8149,
      "step": 786
    },
    {
      "epoch": 0.07020830545519426,
      "grad_norm": 88.5013198852539,
      "learning_rate": 3e-06,
      "loss": -10.8751,
      "step": 787
    },
    {
      "epoch": 0.07029751550024532,
      "grad_norm": 106.22066497802734,
      "learning_rate": 3e-06,
      "loss": -14.8712,
      "step": 788
    },
    {
      "epoch": 0.0703867255452964,
      "grad_norm": 95.8133544921875,
      "learning_rate": 3e-06,
      "loss": -17.8064,
      "step": 789
    },
    {
      "epoch": 0.07047593559034747,
      "grad_norm": 105.98171997070312,
      "learning_rate": 3e-06,
      "loss": -4.9442,
      "step": 790
    },
    {
      "epoch": 0.07056514563539855,
      "grad_norm": 108.74724578857422,
      "learning_rate": 3e-06,
      "loss": -20.1345,
      "step": 791
    },
    {
      "epoch": 0.07065435568044962,
      "grad_norm": 217.0005340576172,
      "learning_rate": 3e-06,
      "loss": -5.632,
      "step": 792
    },
    {
      "completion_length": 137.1041717529297,
      "epoch": 0.0707435657255007,
      "grad_norm": 186.0244140625,
      "learning_rate": 3e-06,
      "loss": 20.4605,
      "reward": 1.8869168162345886,
      "reward_std": 0.7256337702274323,
      "rewards/correctness_reward_func": 1.2916666567325592,
      "rewards/int_reward_func": 0.4583333283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.13691667094826698,
      "step": 793,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.07083277577055176,
      "grad_norm": 139.76405334472656,
      "learning_rate": 3e-06,
      "loss": 14.9958,
      "step": 794
    },
    {
      "epoch": 0.07092198581560284,
      "grad_norm": 155.77352905273438,
      "learning_rate": 3e-06,
      "loss": 24.0522,
      "step": 795
    },
    {
      "epoch": 0.07101119586065391,
      "grad_norm": 147.90013122558594,
      "learning_rate": 3e-06,
      "loss": 30.8391,
      "step": 796
    },
    {
      "epoch": 0.07110040590570499,
      "grad_norm": 154.99143981933594,
      "learning_rate": 3e-06,
      "loss": 24.957,
      "step": 797
    },
    {
      "epoch": 0.07118961595075605,
      "grad_norm": 154.12411499023438,
      "learning_rate": 3e-06,
      "loss": 32.2635,
      "step": 798
    },
    {
      "epoch": 0.07127882599580712,
      "grad_norm": 142.52955627441406,
      "learning_rate": 3e-06,
      "loss": 19.5332,
      "step": 799
    },
    {
      "epoch": 0.0713680360408582,
      "grad_norm": 140.1791229248047,
      "learning_rate": 3e-06,
      "loss": 13.5796,
      "step": 800
    },
    {
      "epoch": 0.07145724608590927,
      "grad_norm": 144.1186981201172,
      "learning_rate": 3e-06,
      "loss": 21.398,
      "step": 801
    },
    {
      "epoch": 0.07154645613096035,
      "grad_norm": 139.25230407714844,
      "learning_rate": 3e-06,
      "loss": 28.6905,
      "step": 802
    },
    {
      "epoch": 0.07163566617601141,
      "grad_norm": 151.95538330078125,
      "learning_rate": 3e-06,
      "loss": 21.6279,
      "step": 803
    },
    {
      "epoch": 0.0717248762210625,
      "grad_norm": 143.84974670410156,
      "learning_rate": 3e-06,
      "loss": 29.0759,
      "step": 804
    },
    {
      "completion_length": 136.22916793823242,
      "epoch": 0.07181408626611356,
      "grad_norm": 92.62751770019531,
      "learning_rate": 3e-06,
      "loss": -21.4484,
      "reward": 2.414271116256714,
      "reward_std": 0.33920496702194214,
      "rewards/correctness_reward_func": 1.8333333134651184,
      "rewards/int_reward_func": 0.4583333283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.12260417267680168,
      "step": 805,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.07190329631116464,
      "grad_norm": 60.42007827758789,
      "learning_rate": 3e-06,
      "loss": -12.7491,
      "step": 806
    },
    {
      "epoch": 0.07199250635621571,
      "grad_norm": 94.1617660522461,
      "learning_rate": 3e-06,
      "loss": -14.5131,
      "step": 807
    },
    {
      "epoch": 0.07208171640126679,
      "grad_norm": 68.65794372558594,
      "learning_rate": 3e-06,
      "loss": -14.1192,
      "step": 808
    },
    {
      "epoch": 0.07217092644631785,
      "grad_norm": 79.4446792602539,
      "learning_rate": 3e-06,
      "loss": -12.4833,
      "step": 809
    },
    {
      "epoch": 0.07226013649136893,
      "grad_norm": 82.45279693603516,
      "learning_rate": 3e-06,
      "loss": -8.9581,
      "step": 810
    },
    {
      "epoch": 0.07234934653642,
      "grad_norm": 101.833984375,
      "learning_rate": 3e-06,
      "loss": -22.4901,
      "step": 811
    },
    {
      "epoch": 0.07243855658147107,
      "grad_norm": 78.60984802246094,
      "learning_rate": 3e-06,
      "loss": -12.9701,
      "step": 812
    },
    {
      "epoch": 0.07252776662652215,
      "grad_norm": 100.63545989990234,
      "learning_rate": 3e-06,
      "loss": -15.4353,
      "step": 813
    },
    {
      "epoch": 0.07261697667157321,
      "grad_norm": 66.36518096923828,
      "learning_rate": 3e-06,
      "loss": -14.7871,
      "step": 814
    },
    {
      "epoch": 0.0727061867166243,
      "grad_norm": 73.31499481201172,
      "learning_rate": 3e-06,
      "loss": -13.5328,
      "step": 815
    },
    {
      "epoch": 0.07279539676167536,
      "grad_norm": 81.8609848022461,
      "learning_rate": 3e-06,
      "loss": -10.202,
      "step": 816
    },
    {
      "completion_length": 176.0416717529297,
      "epoch": 0.07288460680672644,
      "grad_norm": 85.71199035644531,
      "learning_rate": 3e-06,
      "loss": -2.3775,
      "reward": 1.308291733264923,
      "reward_std": 0.4305167943239212,
      "rewards/correctness_reward_func": 0.8333333432674408,
      "rewards/int_reward_func": 0.4583333283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.016624998301267624,
      "step": 817,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.0729738168517775,
      "grad_norm": 113.99066162109375,
      "learning_rate": 3e-06,
      "loss": 3.9682,
      "step": 818
    },
    {
      "epoch": 0.07306302689682859,
      "grad_norm": 90.42432403564453,
      "learning_rate": 3e-06,
      "loss": -8.968,
      "step": 819
    },
    {
      "epoch": 0.07315223694187965,
      "grad_norm": 86.66998291015625,
      "learning_rate": 3e-06,
      "loss": 5.9931,
      "step": 820
    },
    {
      "epoch": 0.07324144698693073,
      "grad_norm": 112.87352752685547,
      "learning_rate": 3e-06,
      "loss": 3.7934,
      "step": 821
    },
    {
      "epoch": 0.0733306570319818,
      "grad_norm": 116.59276580810547,
      "learning_rate": 3e-06,
      "loss": -12.3296,
      "step": 822
    },
    {
      "epoch": 0.07341986707703288,
      "grad_norm": 95.53129577636719,
      "learning_rate": 3e-06,
      "loss": -3.2342,
      "step": 823
    },
    {
      "epoch": 0.07350907712208395,
      "grad_norm": 134.6486053466797,
      "learning_rate": 3e-06,
      "loss": 2.4981,
      "step": 824
    },
    {
      "epoch": 0.07359828716713501,
      "grad_norm": 80.79833221435547,
      "learning_rate": 3e-06,
      "loss": -9.3798,
      "step": 825
    },
    {
      "epoch": 0.07368749721218609,
      "grad_norm": 107.38970184326172,
      "learning_rate": 3e-06,
      "loss": 4.8459,
      "step": 826
    },
    {
      "epoch": 0.07377670725723716,
      "grad_norm": 91.76937866210938,
      "learning_rate": 3e-06,
      "loss": 2.6928,
      "step": 827
    },
    {
      "epoch": 0.07386591730228824,
      "grad_norm": 112.15656280517578,
      "learning_rate": 3e-06,
      "loss": -13.4598,
      "step": 828
    },
    {
      "completion_length": 161.2916717529297,
      "epoch": 0.0739551273473393,
      "grad_norm": 225.7854461669922,
      "learning_rate": 3e-06,
      "loss": -7.2711,
      "reward": 1.8568333387374878,
      "reward_std": 0.38829553686082363,
      "rewards/correctness_reward_func": 1.2916666865348816,
      "rewards/int_reward_func": 0.4895833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.0755833312869072,
      "step": 829,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.07404433739239039,
      "grad_norm": 203.64381408691406,
      "learning_rate": 3e-06,
      "loss": -24.0722,
      "step": 830
    },
    {
      "epoch": 0.07413354743744145,
      "grad_norm": 470.9283752441406,
      "learning_rate": 3e-06,
      "loss": 7.9581,
      "step": 831
    },
    {
      "epoch": 0.07422275748249253,
      "grad_norm": 261.8198547363281,
      "learning_rate": 3e-06,
      "loss": -13.8961,
      "step": 832
    },
    {
      "epoch": 0.0743119675275436,
      "grad_norm": 238.60263061523438,
      "learning_rate": 3e-06,
      "loss": -17.5778,
      "step": 833
    },
    {
      "epoch": 0.07440117757259468,
      "grad_norm": 251.20684814453125,
      "learning_rate": 3e-06,
      "loss": -13.8026,
      "step": 834
    },
    {
      "epoch": 0.07449038761764575,
      "grad_norm": 233.15805053710938,
      "learning_rate": 3e-06,
      "loss": -9.8419,
      "step": 835
    },
    {
      "epoch": 0.07457959766269683,
      "grad_norm": 188.42831420898438,
      "learning_rate": 3e-06,
      "loss": -27.1095,
      "step": 836
    },
    {
      "epoch": 0.07466880770774789,
      "grad_norm": 330.0888671875,
      "learning_rate": 3e-06,
      "loss": 2.663,
      "step": 837
    },
    {
      "epoch": 0.07475801775279897,
      "grad_norm": 187.619873046875,
      "learning_rate": 3e-06,
      "loss": -17.3052,
      "step": 838
    },
    {
      "epoch": 0.07484722779785004,
      "grad_norm": 273.087646484375,
      "learning_rate": 3e-06,
      "loss": -20.1137,
      "step": 839
    },
    {
      "epoch": 0.0749364378429011,
      "grad_norm": 231.94540405273438,
      "learning_rate": 3e-06,
      "loss": -17.6819,
      "step": 840
    },
    {
      "completion_length": 124.5625,
      "epoch": 0.07502564788795218,
      "grad_norm": 169.8292236328125,
      "learning_rate": 3e-06,
      "loss": -14.9169,
      "reward": 2.1858333349227905,
      "reward_std": 0.5357859879732132,
      "rewards/correctness_reward_func": 1.5833333134651184,
      "rewards/int_reward_func": 0.4270833432674408,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.17541665583848953,
      "step": 841,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.07511485793300325,
      "grad_norm": 151.44129943847656,
      "learning_rate": 3e-06,
      "loss": -38.7775,
      "step": 842
    },
    {
      "epoch": 0.07520406797805433,
      "grad_norm": 141.34671020507812,
      "learning_rate": 3e-06,
      "loss": -32.3555,
      "step": 843
    },
    {
      "epoch": 0.0752932780231054,
      "grad_norm": 117.83955383300781,
      "learning_rate": 3e-06,
      "loss": -35.7298,
      "step": 844
    },
    {
      "epoch": 0.07538248806815648,
      "grad_norm": 113.38582611083984,
      "learning_rate": 3e-06,
      "loss": -36.8355,
      "step": 845
    },
    {
      "epoch": 0.07547169811320754,
      "grad_norm": 147.53521728515625,
      "learning_rate": 3e-06,
      "loss": -34.7305,
      "step": 846
    },
    {
      "epoch": 0.07556090815825862,
      "grad_norm": 167.8444061279297,
      "learning_rate": 3e-06,
      "loss": -17.6609,
      "step": 847
    },
    {
      "epoch": 0.07565011820330969,
      "grad_norm": 177.19976806640625,
      "learning_rate": 3e-06,
      "loss": -42.6765,
      "step": 848
    },
    {
      "epoch": 0.07573932824836077,
      "grad_norm": 207.4672393798828,
      "learning_rate": 3e-06,
      "loss": -36.7629,
      "step": 849
    },
    {
      "epoch": 0.07582853829341184,
      "grad_norm": 124.84293365478516,
      "learning_rate": 3e-06,
      "loss": -39.1349,
      "step": 850
    },
    {
      "epoch": 0.07591774833846292,
      "grad_norm": 134.89764404296875,
      "learning_rate": 3e-06,
      "loss": -41.2224,
      "step": 851
    },
    {
      "epoch": 0.07600695838351398,
      "grad_norm": 161.6527862548828,
      "learning_rate": 3e-06,
      "loss": -39.203,
      "step": 852
    },
    {
      "completion_length": 147.7291717529297,
      "epoch": 0.07609616842856505,
      "grad_norm": 371.6798095703125,
      "learning_rate": 3e-06,
      "loss": 88.7083,
      "reward": 2.0986876487731934,
      "reward_std": 0.5909168422222137,
      "rewards/correctness_reward_func": 1.5,
      "rewards/int_reward_func": 0.5,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.09868749976158142,
      "step": 853,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.07618537847361613,
      "grad_norm": 385.34136962890625,
      "learning_rate": 3e-06,
      "loss": 86.0723,
      "step": 854
    },
    {
      "epoch": 0.0762745885186672,
      "grad_norm": 360.78021240234375,
      "learning_rate": 3e-06,
      "loss": 71.1204,
      "step": 855
    },
    {
      "epoch": 0.07636379856371828,
      "grad_norm": 293.267333984375,
      "learning_rate": 3e-06,
      "loss": 66.9494,
      "step": 856
    },
    {
      "epoch": 0.07645300860876934,
      "grad_norm": 440.7154846191406,
      "learning_rate": 3e-06,
      "loss": 88.5771,
      "step": 857
    },
    {
      "epoch": 0.07654221865382042,
      "grad_norm": 327.457275390625,
      "learning_rate": 3e-06,
      "loss": 58.8516,
      "step": 858
    },
    {
      "epoch": 0.07663142869887149,
      "grad_norm": 371.9436340332031,
      "learning_rate": 3e-06,
      "loss": 85.2973,
      "step": 859
    },
    {
      "epoch": 0.07672063874392257,
      "grad_norm": 389.5568542480469,
      "learning_rate": 3e-06,
      "loss": 79.8231,
      "step": 860
    },
    {
      "epoch": 0.07680984878897364,
      "grad_norm": 321.9656066894531,
      "learning_rate": 3e-06,
      "loss": 63.4841,
      "step": 861
    },
    {
      "epoch": 0.07689905883402472,
      "grad_norm": 284.66876220703125,
      "learning_rate": 3e-06,
      "loss": 59.0441,
      "step": 862
    },
    {
      "epoch": 0.07698826887907578,
      "grad_norm": 410.6514587402344,
      "learning_rate": 3e-06,
      "loss": 76.3094,
      "step": 863
    },
    {
      "epoch": 0.07707747892412686,
      "grad_norm": 284.0197448730469,
      "learning_rate": 3e-06,
      "loss": 49.468,
      "step": 864
    },
    {
      "completion_length": 144.1041717529297,
      "epoch": 0.07716668896917793,
      "grad_norm": 226.17822265625,
      "learning_rate": 3e-06,
      "loss": 56.9406,
      "reward": 2.379916787147522,
      "reward_std": 0.430880606174469,
      "rewards/correctness_reward_func": 1.7916666865348816,
      "rewards/int_reward_func": 0.5,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.08824999909847975,
      "step": 865,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.077255899014229,
      "grad_norm": 186.55560302734375,
      "learning_rate": 3e-06,
      "loss": 52.0502,
      "step": 866
    },
    {
      "epoch": 0.07734510905928008,
      "grad_norm": 184.80516052246094,
      "learning_rate": 3e-06,
      "loss": 46.5618,
      "step": 867
    },
    {
      "epoch": 0.07743431910433114,
      "grad_norm": 178.5349884033203,
      "learning_rate": 3e-06,
      "loss": 46.2696,
      "step": 868
    },
    {
      "epoch": 0.07752352914938222,
      "grad_norm": 148.83154296875,
      "learning_rate": 3e-06,
      "loss": 29.464,
      "step": 869
    },
    {
      "epoch": 0.07761273919443329,
      "grad_norm": 161.14889526367188,
      "learning_rate": 3e-06,
      "loss": 51.0483,
      "step": 870
    },
    {
      "epoch": 0.07770194923948437,
      "grad_norm": 192.32308959960938,
      "learning_rate": 3e-06,
      "loss": 47.9675,
      "step": 871
    },
    {
      "epoch": 0.07779115928453544,
      "grad_norm": 152.79583740234375,
      "learning_rate": 3e-06,
      "loss": 43.9552,
      "step": 872
    },
    {
      "epoch": 0.07788036932958652,
      "grad_norm": 151.7612762451172,
      "learning_rate": 3e-06,
      "loss": 38.7329,
      "step": 873
    },
    {
      "epoch": 0.07796957937463758,
      "grad_norm": 133.1282196044922,
      "learning_rate": 3e-06,
      "loss": 38.5692,
      "step": 874
    },
    {
      "epoch": 0.07805878941968866,
      "grad_norm": 103.07962036132812,
      "learning_rate": 3e-06,
      "loss": 24.2915,
      "step": 875
    },
    {
      "epoch": 0.07814799946473973,
      "grad_norm": 129.3807373046875,
      "learning_rate": 3e-06,
      "loss": 43.1067,
      "step": 876
    },
    {
      "completion_length": 127.54166793823242,
      "epoch": 0.07823720950979081,
      "grad_norm": 125.33885192871094,
      "learning_rate": 3e-06,
      "loss": -23.6161,
      "reward": 1.667020857334137,
      "reward_std": 0.5597978234291077,
      "rewards/correctness_reward_func": 1.0833333432674408,
      "rewards/int_reward_func": 0.40625,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.17743750661611557,
      "step": 877,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.07832641955484188,
      "grad_norm": 138.50013732910156,
      "learning_rate": 3e-06,
      "loss": -16.1757,
      "step": 878
    },
    {
      "epoch": 0.07841562959989294,
      "grad_norm": 130.26280212402344,
      "learning_rate": 3e-06,
      "loss": -21.8816,
      "step": 879
    },
    {
      "epoch": 0.07850483964494402,
      "grad_norm": 141.026123046875,
      "learning_rate": 3e-06,
      "loss": -22.0761,
      "step": 880
    },
    {
      "epoch": 0.07859404968999509,
      "grad_norm": 126.53893280029297,
      "learning_rate": 3e-06,
      "loss": -23.4112,
      "step": 881
    },
    {
      "epoch": 0.07868325973504617,
      "grad_norm": 153.45120239257812,
      "learning_rate": 3e-06,
      "loss": -17.7169,
      "step": 882
    },
    {
      "epoch": 0.07877246978009723,
      "grad_norm": 122.84283447265625,
      "learning_rate": 3e-06,
      "loss": -24.8704,
      "step": 883
    },
    {
      "epoch": 0.07886167982514831,
      "grad_norm": 157.95201110839844,
      "learning_rate": 3e-06,
      "loss": -16.2334,
      "step": 884
    },
    {
      "epoch": 0.07895088987019938,
      "grad_norm": 136.01124572753906,
      "learning_rate": 3e-06,
      "loss": -23.3998,
      "step": 885
    },
    {
      "epoch": 0.07904009991525046,
      "grad_norm": 135.98423767089844,
      "learning_rate": 3e-06,
      "loss": -23.6646,
      "step": 886
    },
    {
      "epoch": 0.07912930996030153,
      "grad_norm": 131.24002075195312,
      "learning_rate": 3e-06,
      "loss": -25.7454,
      "step": 887
    },
    {
      "epoch": 0.07921852000535261,
      "grad_norm": 124.26398468017578,
      "learning_rate": 3e-06,
      "loss": -20.446,
      "step": 888
    },
    {
      "completion_length": 153.58333587646484,
      "epoch": 0.07930773005040367,
      "grad_norm": 240.47964477539062,
      "learning_rate": 3e-06,
      "loss": -52.9796,
      "reward": 1.8959583044052124,
      "reward_std": 0.6871494352817535,
      "rewards/correctness_reward_func": 1.3333333432674408,
      "rewards/int_reward_func": 0.4583333283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.10429166257381439,
      "step": 889,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.07939694009545475,
      "grad_norm": 205.4910430908203,
      "learning_rate": 3e-06,
      "loss": -52.7249,
      "step": 890
    },
    {
      "epoch": 0.07948615014050582,
      "grad_norm": 242.2780303955078,
      "learning_rate": 3e-06,
      "loss": -67.3242,
      "step": 891
    },
    {
      "epoch": 0.07957536018555689,
      "grad_norm": 262.0589599609375,
      "learning_rate": 3e-06,
      "loss": -66.9661,
      "step": 892
    },
    {
      "epoch": 0.07966457023060797,
      "grad_norm": 186.11415100097656,
      "learning_rate": 3e-06,
      "loss": -68.4566,
      "step": 893
    },
    {
      "epoch": 0.07975378027565903,
      "grad_norm": 254.95228576660156,
      "learning_rate": 3e-06,
      "loss": -66.0288,
      "step": 894
    },
    {
      "epoch": 0.07984299032071011,
      "grad_norm": 270.2388000488281,
      "learning_rate": 3e-06,
      "loss": -59.1402,
      "step": 895
    },
    {
      "epoch": 0.07993220036576118,
      "grad_norm": 232.1254119873047,
      "learning_rate": 3e-06,
      "loss": -58.3748,
      "step": 896
    },
    {
      "epoch": 0.08002141041081226,
      "grad_norm": 423.2415466308594,
      "learning_rate": 3e-06,
      "loss": -74.5377,
      "step": 897
    },
    {
      "epoch": 0.08011062045586333,
      "grad_norm": 289.6065673828125,
      "learning_rate": 3e-06,
      "loss": -76.1074,
      "step": 898
    },
    {
      "epoch": 0.0801998305009144,
      "grad_norm": 212.4766845703125,
      "learning_rate": 3e-06,
      "loss": -74.2601,
      "step": 899
    },
    {
      "epoch": 0.08028904054596547,
      "grad_norm": 286.6225891113281,
      "learning_rate": 3e-06,
      "loss": -74.3892,
      "step": 900
    },
    {
      "completion_length": 119.62500381469727,
      "epoch": 0.08037825059101655,
      "grad_norm": 123.23504638671875,
      "learning_rate": 3e-06,
      "loss": -22.0192,
      "reward": 2.5415626764297485,
      "reward_std": 0.190566536039114,
      "rewards/correctness_reward_func": 1.9166666865348816,
      "rewards/int_reward_func": 0.4791666716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.1457291655242443,
      "step": 901,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.08046746063606762,
      "grad_norm": 134.62196350097656,
      "learning_rate": 3e-06,
      "loss": -26.971,
      "step": 902
    },
    {
      "epoch": 0.0805566706811187,
      "grad_norm": 211.32479858398438,
      "learning_rate": 3e-06,
      "loss": -31.3724,
      "step": 903
    },
    {
      "epoch": 0.08064588072616977,
      "grad_norm": 160.55055236816406,
      "learning_rate": 3e-06,
      "loss": -22.228,
      "step": 904
    },
    {
      "epoch": 0.08073509077122083,
      "grad_norm": 125.40478515625,
      "learning_rate": 3e-06,
      "loss": -21.2257,
      "step": 905
    },
    {
      "epoch": 0.08082430081627191,
      "grad_norm": 111.1106948852539,
      "learning_rate": 3e-06,
      "loss": -22.3095,
      "step": 906
    },
    {
      "epoch": 0.08091351086132298,
      "grad_norm": 122.1114501953125,
      "learning_rate": 3e-06,
      "loss": -24.7909,
      "step": 907
    },
    {
      "epoch": 0.08100272090637406,
      "grad_norm": 156.01158142089844,
      "learning_rate": 3e-06,
      "loss": -31.0448,
      "step": 908
    },
    {
      "epoch": 0.08109193095142513,
      "grad_norm": 158.0888214111328,
      "learning_rate": 3e-06,
      "loss": -35.1506,
      "step": 909
    },
    {
      "epoch": 0.0811811409964762,
      "grad_norm": 156.11680603027344,
      "learning_rate": 3e-06,
      "loss": -26.2504,
      "step": 910
    },
    {
      "epoch": 0.08127035104152727,
      "grad_norm": 136.36370849609375,
      "learning_rate": 3e-06,
      "loss": -24.5191,
      "step": 911
    },
    {
      "epoch": 0.08135956108657835,
      "grad_norm": 138.4123077392578,
      "learning_rate": 3e-06,
      "loss": -25.2287,
      "step": 912
    },
    {
      "completion_length": 123.97917175292969,
      "epoch": 0.08144877113162942,
      "grad_norm": 69.13970184326172,
      "learning_rate": 3e-06,
      "loss": -4.2059,
      "reward": 2.349874973297119,
      "reward_std": 0.39924251288175583,
      "rewards/correctness_reward_func": 1.75,
      "rewards/int_reward_func": 0.4583333283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.14154166728258133,
      "step": 913,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.0815379811766805,
      "grad_norm": 109.77488708496094,
      "learning_rate": 3e-06,
      "loss": -6.6804,
      "step": 914
    },
    {
      "epoch": 0.08162719122173157,
      "grad_norm": 108.82147216796875,
      "learning_rate": 3e-06,
      "loss": 1.8191,
      "step": 915
    },
    {
      "epoch": 0.08171640126678265,
      "grad_norm": 88.40335083007812,
      "learning_rate": 3e-06,
      "loss": -5.8692,
      "step": 916
    },
    {
      "epoch": 0.08180561131183371,
      "grad_norm": 76.1854019165039,
      "learning_rate": 3e-06,
      "loss": -1.3803,
      "step": 917
    },
    {
      "epoch": 0.08189482135688479,
      "grad_norm": 94.09133911132812,
      "learning_rate": 3e-06,
      "loss": -2.3375,
      "step": 918
    },
    {
      "epoch": 0.08198403140193586,
      "grad_norm": 84.88536071777344,
      "learning_rate": 3e-06,
      "loss": -5.6229,
      "step": 919
    },
    {
      "epoch": 0.08207324144698692,
      "grad_norm": 92.1208267211914,
      "learning_rate": 3e-06,
      "loss": -7.5509,
      "step": 920
    },
    {
      "epoch": 0.082162451492038,
      "grad_norm": 89.02661895751953,
      "learning_rate": 3e-06,
      "loss": 0.5948,
      "step": 921
    },
    {
      "epoch": 0.08225166153708907,
      "grad_norm": 95.09249114990234,
      "learning_rate": 3e-06,
      "loss": -6.4904,
      "step": 922
    },
    {
      "epoch": 0.08234087158214015,
      "grad_norm": 83.8741683959961,
      "learning_rate": 3e-06,
      "loss": -2.875,
      "step": 923
    },
    {
      "epoch": 0.08243008162719122,
      "grad_norm": 129.45420837402344,
      "learning_rate": 3e-06,
      "loss": -3.085,
      "step": 924
    },
    {
      "completion_length": 115.27083587646484,
      "epoch": 0.0825192916722423,
      "grad_norm": 250.4253387451172,
      "learning_rate": 3e-06,
      "loss": 9.4605,
      "reward": 2.1223334074020386,
      "reward_std": 0.7158277630805969,
      "rewards/correctness_reward_func": 1.5,
      "rewards/int_reward_func": 0.4166666716337204,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.205666683614254,
      "step": 925,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.08260850171729336,
      "grad_norm": 318.8369140625,
      "learning_rate": 3e-06,
      "loss": 16.4358,
      "step": 926
    },
    {
      "epoch": 0.08269771176234444,
      "grad_norm": 314.372314453125,
      "learning_rate": 3e-06,
      "loss": 5.4103,
      "step": 927
    },
    {
      "epoch": 0.08278692180739551,
      "grad_norm": 255.00933837890625,
      "learning_rate": 3e-06,
      "loss": -3.1256,
      "step": 928
    },
    {
      "epoch": 0.08287613185244659,
      "grad_norm": 357.3619384765625,
      "learning_rate": 3e-06,
      "loss": 3.4132,
      "step": 929
    },
    {
      "epoch": 0.08296534189749766,
      "grad_norm": 409.3254089355469,
      "learning_rate": 3e-06,
      "loss": 23.0602,
      "step": 930
    },
    {
      "epoch": 0.08305455194254874,
      "grad_norm": 270.6861877441406,
      "learning_rate": 3e-06,
      "loss": 8.4527,
      "step": 931
    },
    {
      "epoch": 0.0831437619875998,
      "grad_norm": 507.520263671875,
      "learning_rate": 3e-06,
      "loss": 14.712,
      "step": 932
    },
    {
      "epoch": 0.08323297203265087,
      "grad_norm": 281.0194091796875,
      "learning_rate": 3e-06,
      "loss": 4.5989,
      "step": 933
    },
    {
      "epoch": 0.08332218207770195,
      "grad_norm": 275.3479309082031,
      "learning_rate": 3e-06,
      "loss": -5.1609,
      "step": 934
    },
    {
      "epoch": 0.08341139212275302,
      "grad_norm": 358.3206481933594,
      "learning_rate": 3e-06,
      "loss": 3.3467,
      "step": 935
    },
    {
      "epoch": 0.0835006021678041,
      "grad_norm": 403.45440673828125,
      "learning_rate": 3e-06,
      "loss": 20.7584,
      "step": 936
    },
    {
      "completion_length": 131.7291717529297,
      "epoch": 0.08358981221285516,
      "grad_norm": 112.89104461669922,
      "learning_rate": 3e-06,
      "loss": -12.4613,
      "reward": 1.9657083749771118,
      "reward_std": 0.33454202115535736,
      "rewards/correctness_reward_func": 1.3333333432674408,
      "rewards/int_reward_func": 0.5,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.13237499818205833,
      "step": 937,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.08367902225790624,
      "grad_norm": 158.0906524658203,
      "learning_rate": 3e-06,
      "loss": -6.7935,
      "step": 938
    },
    {
      "epoch": 0.08376823230295731,
      "grad_norm": 127.69352722167969,
      "learning_rate": 3e-06,
      "loss": -15.3734,
      "step": 939
    },
    {
      "epoch": 0.08385744234800839,
      "grad_norm": 207.05262756347656,
      "learning_rate": 3e-06,
      "loss": -25.0891,
      "step": 940
    },
    {
      "epoch": 0.08394665239305946,
      "grad_norm": 546.4678344726562,
      "learning_rate": 3e-06,
      "loss": -29.0194,
      "step": 941
    },
    {
      "epoch": 0.08403586243811054,
      "grad_norm": 141.02198791503906,
      "learning_rate": 3e-06,
      "loss": -10.7265,
      "step": 942
    },
    {
      "epoch": 0.0841250724831616,
      "grad_norm": 137.6843719482422,
      "learning_rate": 3e-06,
      "loss": -13.3029,
      "step": 943
    },
    {
      "epoch": 0.08421428252821268,
      "grad_norm": 211.74227905273438,
      "learning_rate": 3e-06,
      "loss": -8.8958,
      "step": 944
    },
    {
      "epoch": 0.08430349257326375,
      "grad_norm": 123.87110900878906,
      "learning_rate": 3e-06,
      "loss": -16.9913,
      "step": 945
    },
    {
      "epoch": 0.08439270261831482,
      "grad_norm": 206.8551025390625,
      "learning_rate": 3e-06,
      "loss": -26.9321,
      "step": 946
    },
    {
      "epoch": 0.0844819126633659,
      "grad_norm": 193.33346557617188,
      "learning_rate": 3e-06,
      "loss": -30.975,
      "step": 947
    },
    {
      "epoch": 0.08457112270841696,
      "grad_norm": 147.73297119140625,
      "learning_rate": 3e-06,
      "loss": -13.515,
      "step": 948
    },
    {
      "completion_length": 159.95833587646484,
      "epoch": 0.08466033275346804,
      "grad_norm": 186.3380889892578,
      "learning_rate": 3e-06,
      "loss": 9.8972,
      "reward": 1.8234166502952576,
      "reward_std": 0.42612800002098083,
      "rewards/correctness_reward_func": 1.2916666865348816,
      "rewards/int_reward_func": 0.4375,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.09425000101327896,
      "step": 949,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.08474954279851911,
      "grad_norm": 178.0104522705078,
      "learning_rate": 3e-06,
      "loss": 10.3947,
      "step": 950
    },
    {
      "epoch": 0.08483875284357019,
      "grad_norm": 160.924560546875,
      "learning_rate": 3e-06,
      "loss": 8.1323,
      "step": 951
    },
    {
      "epoch": 0.08492796288862126,
      "grad_norm": 144.38978576660156,
      "learning_rate": 3e-06,
      "loss": 5.1531,
      "step": 952
    },
    {
      "epoch": 0.08501717293367234,
      "grad_norm": 174.2298126220703,
      "learning_rate": 3e-06,
      "loss": 5.4548,
      "step": 953
    },
    {
      "epoch": 0.0851063829787234,
      "grad_norm": 164.93479919433594,
      "learning_rate": 3e-06,
      "loss": 10.7603,
      "step": 954
    },
    {
      "epoch": 0.08519559302377448,
      "grad_norm": 198.3860626220703,
      "learning_rate": 3e-06,
      "loss": 9.1507,
      "step": 955
    },
    {
      "epoch": 0.08528480306882555,
      "grad_norm": 160.76519775390625,
      "learning_rate": 3e-06,
      "loss": 9.4591,
      "step": 956
    },
    {
      "epoch": 0.08537401311387663,
      "grad_norm": 170.39776611328125,
      "learning_rate": 3e-06,
      "loss": 7.0709,
      "step": 957
    },
    {
      "epoch": 0.0854632231589277,
      "grad_norm": 145.32798767089844,
      "learning_rate": 3e-06,
      "loss": 4.1507,
      "step": 958
    },
    {
      "epoch": 0.08555243320397876,
      "grad_norm": 170.50514221191406,
      "learning_rate": 3e-06,
      "loss": 4.4292,
      "step": 959
    },
    {
      "epoch": 0.08564164324902984,
      "grad_norm": 197.32290649414062,
      "learning_rate": 3e-06,
      "loss": 10.1245,
      "step": 960
    },
    {
      "completion_length": 149.68750762939453,
      "epoch": 0.08573085329408091,
      "grad_norm": 530.301025390625,
      "learning_rate": 3e-06,
      "loss": 6.2918,
      "reward": 2.106416702270508,
      "reward_std": 0.5645134299993515,
      "rewards/correctness_reward_func": 1.5416666865348816,
      "rewards/int_reward_func": 0.4791666567325592,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.08558332687243819,
      "step": 961,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.08582006333913199,
      "grad_norm": 593.3743286132812,
      "learning_rate": 3e-06,
      "loss": -27.7628,
      "step": 962
    },
    {
      "epoch": 0.08590927338418305,
      "grad_norm": 378.6949157714844,
      "learning_rate": 3e-06,
      "loss": -25.6701,
      "step": 963
    },
    {
      "epoch": 0.08599848342923413,
      "grad_norm": 402.416748046875,
      "learning_rate": 3e-06,
      "loss": 10.4982,
      "step": 964
    },
    {
      "epoch": 0.0860876934742852,
      "grad_norm": 443.3346862792969,
      "learning_rate": 3e-06,
      "loss": -39.2448,
      "step": 965
    },
    {
      "epoch": 0.08617690351933628,
      "grad_norm": 401.20574951171875,
      "learning_rate": 3e-06,
      "loss": -68.5925,
      "step": 966
    },
    {
      "epoch": 0.08626611356438735,
      "grad_norm": 553.38720703125,
      "learning_rate": 3e-06,
      "loss": 1.8602,
      "step": 967
    },
    {
      "epoch": 0.08635532360943843,
      "grad_norm": 628.0134887695312,
      "learning_rate": 3e-06,
      "loss": -36.1923,
      "step": 968
    },
    {
      "epoch": 0.0864445336544895,
      "grad_norm": 380.9430847167969,
      "learning_rate": 3e-06,
      "loss": -32.3079,
      "step": 969
    },
    {
      "epoch": 0.08653374369954057,
      "grad_norm": 385.8163146972656,
      "learning_rate": 3e-06,
      "loss": 3.3819,
      "step": 970
    },
    {
      "epoch": 0.08662295374459164,
      "grad_norm": 432.78118896484375,
      "learning_rate": 3e-06,
      "loss": -46.8594,
      "step": 971
    },
    {
      "epoch": 0.08671216378964271,
      "grad_norm": 439.5821533203125,
      "learning_rate": 3e-06,
      "loss": -77.4002,
      "step": 972
    },
    {
      "completion_length": 121.68750381469727,
      "epoch": 0.08680137383469379,
      "grad_norm": 35.27005386352539,
      "learning_rate": 3e-06,
      "loss": 3.8068,
      "reward": 2.284437596797943,
      "reward_std": 0.15820645913481712,
      "rewards/correctness_reward_func": 1.625,
      "rewards/int_reward_func": 0.5,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.15943749248981476,
      "step": 973,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.08689058387974485,
      "grad_norm": 68.60123443603516,
      "learning_rate": 3e-06,
      "loss": 5.8697,
      "step": 974
    },
    {
      "epoch": 0.08697979392479593,
      "grad_norm": 69.07678985595703,
      "learning_rate": 3e-06,
      "loss": 2.0005,
      "step": 975
    },
    {
      "epoch": 0.087069003969847,
      "grad_norm": 39.30900955200195,
      "learning_rate": 3e-06,
      "loss": 6.0659,
      "step": 976
    },
    {
      "epoch": 0.08715821401489808,
      "grad_norm": 77.51853942871094,
      "learning_rate": 3e-06,
      "loss": 5.9335,
      "step": 977
    },
    {
      "epoch": 0.08724742405994915,
      "grad_norm": 60.07703399658203,
      "learning_rate": 3e-06,
      "loss": 3.043,
      "step": 978
    },
    {
      "epoch": 0.08733663410500023,
      "grad_norm": 39.25843811035156,
      "learning_rate": 3e-06,
      "loss": 3.6489,
      "step": 979
    },
    {
      "epoch": 0.0874258441500513,
      "grad_norm": 46.68893051147461,
      "learning_rate": 3e-06,
      "loss": 5.0957,
      "step": 980
    },
    {
      "epoch": 0.08751505419510237,
      "grad_norm": 55.4852180480957,
      "learning_rate": 3e-06,
      "loss": 0.9956,
      "step": 981
    },
    {
      "epoch": 0.08760426424015344,
      "grad_norm": 51.21168518066406,
      "learning_rate": 3e-06,
      "loss": 5.5624,
      "step": 982
    },
    {
      "epoch": 0.08769347428520452,
      "grad_norm": 64.15937805175781,
      "learning_rate": 3e-06,
      "loss": 4.0323,
      "step": 983
    },
    {
      "epoch": 0.08778268433025559,
      "grad_norm": 65.25579833984375,
      "learning_rate": 3e-06,
      "loss": 1.7638,
      "step": 984
    },
    {
      "completion_length": 138.68750762939453,
      "epoch": 0.08787189437530665,
      "grad_norm": 277.1241760253906,
      "learning_rate": 3e-06,
      "loss": 10.9365,
      "reward": 2.176750063896179,
      "reward_std": 0.40325865149497986,
      "rewards/correctness_reward_func": 1.5833333134651184,
      "rewards/int_reward_func": 0.4791666567325592,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.11425000056624413,
      "step": 985,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.08796110442035773,
      "grad_norm": 363.5508117675781,
      "learning_rate": 3e-06,
      "loss": -2.6393,
      "step": 986
    },
    {
      "epoch": 0.0880503144654088,
      "grad_norm": 337.4767150878906,
      "learning_rate": 3e-06,
      "loss": 9.4007,
      "step": 987
    },
    {
      "epoch": 0.08813952451045988,
      "grad_norm": 292.9395751953125,
      "learning_rate": 3e-06,
      "loss": -2.1309,
      "step": 988
    },
    {
      "epoch": 0.08822873455551095,
      "grad_norm": 246.76112365722656,
      "learning_rate": 3e-06,
      "loss": -9.132,
      "step": 989
    },
    {
      "epoch": 0.08831794460056203,
      "grad_norm": 267.3565368652344,
      "learning_rate": 3e-06,
      "loss": -11.7626,
      "step": 990
    },
    {
      "epoch": 0.08840715464561309,
      "grad_norm": 257.4312438964844,
      "learning_rate": 3e-06,
      "loss": 5.7132,
      "step": 991
    },
    {
      "epoch": 0.08849636469066417,
      "grad_norm": 317.8547058105469,
      "learning_rate": 3e-06,
      "loss": -9.7682,
      "step": 992
    },
    {
      "epoch": 0.08858557473571524,
      "grad_norm": 260.3039855957031,
      "learning_rate": 3e-06,
      "loss": 2.4633,
      "step": 993
    },
    {
      "epoch": 0.08867478478076632,
      "grad_norm": 261.14697265625,
      "learning_rate": 3e-06,
      "loss": -9.4669,
      "step": 994
    },
    {
      "epoch": 0.08876399482581739,
      "grad_norm": 181.8609161376953,
      "learning_rate": 3e-06,
      "loss": -13.1009,
      "step": 995
    },
    {
      "epoch": 0.08885320487086847,
      "grad_norm": 236.33563232421875,
      "learning_rate": 3e-06,
      "loss": -15.7579,
      "step": 996
    },
    {
      "completion_length": 168.9791717529297,
      "epoch": 0.08894241491591953,
      "grad_norm": 502.4021911621094,
      "learning_rate": 3e-06,
      "loss": -65.14,
      "reward": 1.9697707891464233,
      "reward_std": 0.505499929189682,
      "rewards/correctness_reward_func": 1.4166666269302368,
      "rewards/int_reward_func": 0.4895833283662796,
      "rewards/soft_format_reward_func": 0.0,
      "rewards/strict_format_reward_func": 0.0,
      "rewards/xmlcount_reward_func": 0.06352083757519722,
      "step": 997,
      "zero_std_ratio": 0.0
    },
    {
      "epoch": 0.0890316249609706,
      "grad_norm": 564.5300903320312,
      "learning_rate": 3e-06,
      "loss": -45.4159,
      "step": 998
    },
    {
      "epoch": 0.08912083500602168,
      "grad_norm": 453.83160400390625,
      "learning_rate": 3e-06,
      "loss": -31.2339,
      "step": 999
    },
    {
      "epoch": 0.08921004505107274,
      "grad_norm": 449.2571716308594,
      "learning_rate": 3e-06,
      "loss": -71.5707,
      "step": 1000
    }
  ],
  "logging_steps": 1,
  "max_steps": 112090,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": true,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}