{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9850746268656716,
  "eval_steps": 100,
  "global_step": 132,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 823.762321472168,
      "epoch": 0.014925373134328358,
      "grad_norm": 0.17830021679401398,
      "learning_rate": 7.142857142857142e-08,
      "loss": 0.0919,
      "num_tokens": 865267.0,
      "reward": 2.096768334507942,
      "reward_std": 0.32721264474093914,
      "rewards/accuracy_reward": 0.7645089328289032,
      "rewards/format_reward": 0.9899553433060646,
      "rewards/log_scaled_reward": 0.3423039447516203,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 465.7957763671875,
      "epoch": 0.029850746268656716,
      "grad_norm": 0.5716150403022766,
      "learning_rate": 1.4285714285714285e-07,
      "loss": 0.1518,
      "num_tokens": 1414092.0,
      "reward": 0.1841081934981048,
      "reward_std": 0.6422095634043217,
      "rewards/accuracy_reward": 0.23437499813735485,
      "rewards/format_reward": 0.05022321571595967,
      "rewards/log_scaled_reward": -0.10049002850428224,
      "step": 2
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 482.07144927978516,
      "epoch": 0.04477611940298507,
      "grad_norm": 0.532157301902771,
      "learning_rate": 2.1428571428571426e-07,
      "loss": 0.1335,
      "num_tokens": 1992932.0,
      "reward": 0.10916334297508001,
      "reward_std": 0.6189606413245201,
      "rewards/accuracy_reward": 0.20647321455180645,
      "rewards/format_reward": 0.053571428754366934,
      "rewards/log_scaled_reward": -0.1508813016116619,
      "step": 3
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 529.5971221923828,
      "epoch": 0.05970149253731343,
      "grad_norm": 0.42384153604507446,
      "learning_rate": 2.857142857142857e-07,
      "loss": 0.2014,
      "num_tokens": 2596963.0,
      "reward": 0.05200400925241411,
      "reward_std": 0.582377951592207,
      "rewards/accuracy_reward": 0.18191964272409678,
      "rewards/format_reward": 0.04464285809081048,
      "rewards/log_scaled_reward": -0.17455849051475525,
      "step": 4
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 490.2690010070801,
      "epoch": 0.07462686567164178,
      "grad_norm": 0.42382651567459106,
      "learning_rate": 3.5714285714285716e-07,
      "loss": 0.1441,
      "num_tokens": 3173748.0,
      "reward": 0.10420671524479985,
      "reward_std": 0.5982032977044582,
      "rewards/accuracy_reward": 0.18750000279396772,
      "rewards/format_reward": 0.07366071455180645,
      "rewards/log_scaled_reward": -0.15695400722324848,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 450.69086837768555,
      "epoch": 0.08955223880597014,
      "grad_norm": 0.5930183529853821,
      "learning_rate": 4.285714285714285e-07,
      "loss": 0.12,
      "num_tokens": 3702055.0,
      "reward": 0.16118104895576835,
      "reward_std": 0.6403544321656227,
      "rewards/accuracy_reward": 0.20982143096625805,
      "rewards/format_reward": 0.07142857182770967,
      "rewards/log_scaled_reward": -0.12006895546801388,
      "step": 6
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 469.174129486084,
      "epoch": 0.1044776119402985,
      "grad_norm": 0.5980175137519836,
      "learning_rate": 5e-07,
      "loss": 0.1409,
      "num_tokens": 4251483.0,
      "reward": 0.18134657479822636,
      "reward_std": 0.680236354470253,
      "rewards/accuracy_reward": 0.20982142724096775,
      "rewards/format_reward": 0.09821428637951612,
      "rewards/log_scaled_reward": -0.12668914068490267,
      "step": 7
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 501.63505935668945,
      "epoch": 0.11940298507462686,
      "grad_norm": 0.7299549579620361,
      "learning_rate": 5.714285714285714e-07,
      "loss": 0.1195,
      "num_tokens": 4818596.0,
      "reward": 0.20046980120241642,
      "reward_std": 0.67672199010849,
      "rewards/accuracy_reward": 0.2120535708963871,
      "rewards/format_reward": 0.1183035708963871,
      "rewards/log_scaled_reward": -0.1298873471096158,
      "step": 8
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 495.338191986084,
      "epoch": 0.13432835820895522,
      "grad_norm": 0.4346173405647278,
      "learning_rate": 6.428571428571429e-07,
      "loss": 0.1516,
      "num_tokens": 5387755.0,
      "reward": 0.2097643855959177,
      "reward_std": 0.6800287291407585,
      "rewards/accuracy_reward": 0.2220982164144516,
      "rewards/format_reward": 0.10937500093132257,
      "rewards/log_scaled_reward": -0.12170884059742093,
      "step": 9
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 477.33930587768555,
      "epoch": 0.14925373134328357,
      "grad_norm": 3.780118942260742,
      "learning_rate": 7.142857142857143e-07,
      "loss": 0.0884,
      "num_tokens": 5940091.0,
      "reward": 0.31957440078258514,
      "reward_std": 0.7821567356586456,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.16183035727590322,
      "rewards/log_scaled_reward": -0.092255964060314,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 529.9196739196777,
      "epoch": 0.16417910447761194,
      "grad_norm": 0.7230793833732605,
      "learning_rate": 7.857142857142856e-07,
      "loss": 0.0843,
      "num_tokens": 6544955.0,
      "reward": 0.2673153153154999,
      "reward_std": 0.7341821119189262,
      "rewards/accuracy_reward": 0.21316964086145163,
      "rewards/format_reward": 0.1897321417927742,
      "rewards/log_scaled_reward": -0.13558648666366935,
      "step": 11
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 498.72323989868164,
      "epoch": 0.1791044776119403,
      "grad_norm": 0.9150818586349487,
      "learning_rate": 8.57142857142857e-07,
      "loss": 0.1105,
      "num_tokens": 7121891.0,
      "reward": 0.34613738395273685,
      "reward_std": 0.7445657253265381,
      "rewards/accuracy_reward": 0.2388392873108387,
      "rewards/format_reward": 0.22544642630964518,
      "rewards/log_scaled_reward": -0.11814834456890821,
      "step": 12
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 483.24221420288086,
      "epoch": 0.19402985074626866,
      "grad_norm": 2.0404834747314453,
      "learning_rate": 9.285714285714285e-07,
      "loss": 0.0535,
      "num_tokens": 7678060.0,
      "reward": 0.5936555862426758,
      "reward_std": 0.7919039279222488,
      "rewards/accuracy_reward": 0.2667410708963871,
      "rewards/format_reward": 0.4084821417927742,
      "rewards/log_scaled_reward": -0.08156766439788043,
      "step": 13
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 416.2042541503906,
      "epoch": 0.208955223880597,
      "grad_norm": 3.1881885528564453,
      "learning_rate": 1e-06,
      "loss": 0.0238,
      "num_tokens": 8174851.0,
      "reward": 0.6864497661590576,
      "reward_std": 0.8334432542324066,
      "rewards/accuracy_reward": 0.2354910708963871,
      "rewards/format_reward": 0.5345982126891613,
      "rewards/log_scaled_reward": -0.0836395358783193,
      "step": 14
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 402.8460006713867,
      "epoch": 0.22388059701492538,
      "grad_norm": 1.9057585000991821,
      "learning_rate": 9.998286624877785e-07,
      "loss": 0.0362,
      "num_tokens": 8650305.0,
      "reward": 0.7661240547895432,
      "reward_std": 0.8178009614348412,
      "rewards/accuracy_reward": 0.22656249813735485,
      "rewards/format_reward": 0.6272321343421936,
      "rewards/log_scaled_reward": -0.08767060440732166,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 452.4196662902832,
      "epoch": 0.23880597014925373,
      "grad_norm": 25.345657348632812,
      "learning_rate": 9.99314767377287e-07,
      "loss": 0.0285,
      "num_tokens": 9179041.0,
      "reward": 0.875191293656826,
      "reward_std": 0.7929042428731918,
      "rewards/accuracy_reward": 0.25781249813735485,
      "rewards/format_reward": 0.6964285671710968,
      "rewards/log_scaled_reward": -0.07904981379397213,
      "step": 16
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 416.35605239868164,
      "epoch": 0.2537313432835821,
      "grad_norm": 1.1184340715408325,
      "learning_rate": 9.98458666866564e-07,
      "loss": 0.0563,
      "num_tokens": 9701832.0,
      "reward": 0.9071941375732422,
      "reward_std": 0.8025857880711555,
      "rewards/accuracy_reward": 0.24999999813735485,
      "rewards/format_reward": 0.7265625,
      "rewards/log_scaled_reward": -0.06936839601257816,
      "step": 17
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 415.1797065734863,
      "epoch": 0.26865671641791045,
      "grad_norm": 0.4797329604625702,
      "learning_rate": 9.972609476841365e-07,
      "loss": 0.1162,
      "num_tokens": 10210017.0,
      "reward": 0.9835792705416679,
      "reward_std": 0.7636988162994385,
      "rewards/accuracy_reward": 0.24665178544819355,
      "rewards/format_reward": 0.8147321343421936,
      "rewards/log_scaled_reward": -0.07780469593126327,
      "step": 18
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 441.59042739868164,
      "epoch": 0.2835820895522388,
      "grad_norm": 0.4748988151550293,
      "learning_rate": 9.957224306869053e-07,
      "loss": 0.0578,
      "num_tokens": 10730474.0,
      "reward": 1.0904420465230942,
      "reward_std": 0.80109953135252,
      "rewards/accuracy_reward": 0.300223208963871,
      "rewards/format_reward": 0.8158482164144516,
      "rewards/log_scaled_reward": -0.025629449490224943,
      "step": 19
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 426.41743087768555,
      "epoch": 0.29850746268656714,
      "grad_norm": 0.47011756896972656,
      "learning_rate": 9.938441702975689e-07,
      "loss": 0.0503,
      "num_tokens": 11239824.0,
      "reward": 1.2500263825058937,
      "reward_std": 0.8551982864737511,
      "rewards/accuracy_reward": 0.3671875,
      "rewards/format_reward": 0.8526785746216774,
      "rewards/log_scaled_reward": 0.030160245776642114,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 459.67078399658203,
      "epoch": 0.31343283582089554,
      "grad_norm": 0.37291309237480164,
      "learning_rate": 9.916274537819773e-07,
      "loss": 0.0366,
      "num_tokens": 11776161.0,
      "reward": 1.3390378654003143,
      "reward_std": 0.8277674093842506,
      "rewards/accuracy_reward": 0.4196428544819355,
      "rewards/format_reward": 0.8526785746216774,
      "rewards/log_scaled_reward": 0.06671636505052447,
      "step": 21
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 416.93639755249023,
      "epoch": 0.3283582089552239,
      "grad_norm": 0.4895838499069214,
      "learning_rate": 9.890738003669027e-07,
      "loss": 0.0473,
      "num_tokens": 12279200.0,
      "reward": 1.5237962007522583,
      "reward_std": 0.8290813863277435,
      "rewards/accuracy_reward": 0.4810267835855484,
      "rewards/format_reward": 0.8950892835855484,
      "rewards/log_scaled_reward": 0.14768002880737185,
      "step": 22
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 409.1294860839844,
      "epoch": 0.34328358208955223,
      "grad_norm": 0.4449058175086975,
      "learning_rate": 9.861849601988383e-07,
      "loss": 0.0255,
      "num_tokens": 12776356.0,
      "reward": 1.5605345666408539,
      "reward_std": 0.8202421888709068,
      "rewards/accuracy_reward": 0.488839291036129,
      "rewards/format_reward": 0.9151785746216774,
      "rewards/log_scaled_reward": 0.15651662228628993,
      "step": 23
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 487.7277069091797,
      "epoch": 0.3582089552238806,
      "grad_norm": 0.33160600066185,
      "learning_rate": 9.82962913144534e-07,
      "loss": 0.0846,
      "num_tokens": 13349480.0,
      "reward": 1.634689912199974,
      "reward_std": 0.7854569926857948,
      "rewards/accuracy_reward": 0.5301339291036129,
      "rewards/format_reward": 0.9151785746216774,
      "rewards/log_scaled_reward": 0.18937731813639402,
      "step": 24
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 451.1272506713867,
      "epoch": 0.373134328358209,
      "grad_norm": 0.3235984444618225,
      "learning_rate": 9.794098674340966e-07,
      "loss": 0.0424,
      "num_tokens": 13868850.0,
      "reward": 1.9291264861822128,
      "reward_std": 0.6902804151177406,
      "rewards/accuracy_reward": 0.6618303433060646,
      "rewards/format_reward": 0.9441964328289032,
      "rewards/log_scaled_reward": 0.32309958525002,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 516.2678871154785,
      "epoch": 0.3880597014925373,
      "grad_norm": 2.3632936477661133,
      "learning_rate": 9.755282581475767e-07,
      "loss": 0.1035,
      "num_tokens": 14469026.0,
      "reward": 1.7230691313743591,
      "reward_std": 0.6944096386432648,
      "rewards/accuracy_reward": 0.5714285708963871,
      "rewards/format_reward": 0.9285714253783226,
      "rewards/log_scaled_reward": 0.22306904755532742,
      "step": 26
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 504.8772506713867,
      "epoch": 0.40298507462686567,
      "grad_norm": 0.28064557909965515,
      "learning_rate": 9.713207455460892e-07,
      "loss": 0.0575,
      "num_tokens": 15048084.0,
      "reward": 1.8045607656240463,
      "reward_std": 0.6639576852321625,
      "rewards/accuracy_reward": 0.6127232238650322,
      "rewards/format_reward": 0.9274553582072258,
      "rewards/log_scaled_reward": 0.2643821220844984,
      "step": 27
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 491.07703018188477,
      "epoch": 0.417910447761194,
      "grad_norm": 0.29867812991142273,
      "learning_rate": 9.667902132486008e-07,
      "loss": 0.0641,
      "num_tokens": 15607481.0,
      "reward": 1.9283190667629242,
      "reward_std": 0.627920113503933,
      "rewards/accuracy_reward": 0.6529017835855484,
      "rewards/format_reward": 0.9575892761349678,
      "rewards/log_scaled_reward": 0.3178279069252312,
      "step": 28
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 559.9810523986816,
      "epoch": 0.43283582089552236,
      "grad_norm": 0.2834532558917999,
      "learning_rate": 9.619397662556433e-07,
      "loss": 0.0735,
      "num_tokens": 16248464.0,
      "reward": 1.7917230874300003,
      "reward_std": 0.527396660298109,
      "rewards/accuracy_reward": 0.6037946455180645,
      "rewards/format_reward": 0.954241082072258,
      "rewards/log_scaled_reward": 0.23368733504321426,
      "step": 29
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 510.0111885070801,
      "epoch": 0.44776119402985076,
      "grad_norm": 0.25292134284973145,
      "learning_rate": 9.567727288213004e-07,
      "loss": 0.0778,
      "num_tokens": 16828154.0,
      "reward": 2.033374920487404,
      "reward_std": 0.5300325341522694,
      "rewards/accuracy_reward": 0.7008928507566452,
      "rewards/format_reward": 0.967633917927742,
      "rewards/log_scaled_reward": 0.36484804935753345,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 505.8872985839844,
      "epoch": 0.4626865671641791,
      "grad_norm": 0.2629943788051605,
      "learning_rate": 9.512926421749303e-07,
      "loss": 0.0917,
      "num_tokens": 17405221.0,
      "reward": 1.9693890661001205,
      "reward_std": 0.4761252626776695,
      "rewards/accuracy_reward": 0.6595982164144516,
      "rewards/format_reward": 0.9743303507566452,
      "rewards/log_scaled_reward": 0.33546042814850807,
      "step": 31
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 468.31809997558594,
      "epoch": 0.47761194029850745,
      "grad_norm": 0.25987353920936584,
      "learning_rate": 9.455032620941839e-07,
      "loss": 0.116,
      "num_tokens": 17953570.0,
      "reward": 2.0616614371538162,
      "reward_std": 0.49665234982967377,
      "rewards/accuracy_reward": 0.7075892761349678,
      "rewards/format_reward": 0.962053582072258,
      "rewards/log_scaled_reward": 0.39201846718788147,
      "step": 32
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 504.4788246154785,
      "epoch": 0.4925373134328358,
      "grad_norm": 0.27474886178970337,
      "learning_rate": 9.394085563309826e-07,
      "loss": 0.1112,
      "num_tokens": 18531095.0,
      "reward": 1.9522841572761536,
      "reward_std": 0.5375584103167057,
      "rewards/accuracy_reward": 0.6607142835855484,
      "rewards/format_reward": 0.965401791036129,
      "rewards/log_scaled_reward": 0.32616803981363773,
      "step": 33
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 498.3616371154785,
      "epoch": 0.5074626865671642,
      "grad_norm": 2.524815320968628,
      "learning_rate": 9.330127018922193e-07,
      "loss": 0.1055,
      "num_tokens": 19108059.0,
      "reward": 1.9061091989278793,
      "reward_std": 0.5573387667536736,
      "rewards/accuracy_reward": 0.6339285708963871,
      "rewards/format_reward": 0.9642857164144516,
      "rewards/log_scaled_reward": 0.30789486039429903,
      "step": 34
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 444.20873260498047,
      "epoch": 0.5223880597014925,
      "grad_norm": 0.3097524046897888,
      "learning_rate": 9.26320082177046e-07,
      "loss": 0.1342,
      "num_tokens": 19646062.0,
      "reward": 1.9290964603424072,
      "reward_std": 0.5330292023718357,
      "rewards/accuracy_reward": 0.6272321492433548,
      "rewards/format_reward": 0.9765625,
      "rewards/log_scaled_reward": 0.325301731005311,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 420.0848388671875,
      "epoch": 0.5373134328358209,
      "grad_norm": 0.3493054211139679,
      "learning_rate": 9.19335283972712e-07,
      "loss": 0.1097,
      "num_tokens": 20165994.0,
      "reward": 1.937290906906128,
      "reward_std": 0.5374783836305141,
      "rewards/accuracy_reward": 0.621651791036129,
      "rewards/format_reward": 0.9732142835855484,
      "rewards/log_scaled_reward": 0.34242471773177385,
      "step": 36
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 367.64064025878906,
      "epoch": 0.5522388059701493,
      "grad_norm": 0.39517220854759216,
      "learning_rate": 9.120630943110077e-07,
      "loss": 0.1309,
      "num_tokens": 20622824.0,
      "reward": 1.9518826305866241,
      "reward_std": 0.5057090371847153,
      "rewards/accuracy_reward": 0.6294642947614193,
      "rewards/format_reward": 0.9765625074505806,
      "rewards/log_scaled_reward": 0.34585576388053596,
      "step": 37
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 344.1038112640381,
      "epoch": 0.5671641791044776,
      "grad_norm": 1.3338432312011719,
      "learning_rate": 9.045084971874737e-07,
      "loss": 0.1333,
      "num_tokens": 21076925.0,
      "reward": 1.9407142996788025,
      "reward_std": 0.5814780332148075,
      "rewards/accuracy_reward": 0.6127232126891613,
      "rewards/format_reward": 0.9709821417927742,
      "rewards/log_scaled_reward": 0.35700881760567427,
      "step": 38
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 277.19309425354004,
      "epoch": 0.582089552238806,
      "grad_norm": 0.4697703421115875,
      "learning_rate": 8.966766701456176e-07,
      "loss": 0.1089,
      "num_tokens": 21451954.0,
      "reward": 1.8939976394176483,
      "reward_std": 0.531686820089817,
      "rewards/accuracy_reward": 0.5703125037252903,
      "rewards/format_reward": 0.984375,
      "rewards/log_scaled_reward": 0.33931003510951996,
      "step": 39
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 292.2109489440918,
      "epoch": 0.5970149253731343,
      "grad_norm": 0.42749401926994324,
      "learning_rate": 8.885729807284854e-07,
      "loss": 0.1051,
      "num_tokens": 21837471.0,
      "reward": 1.7810039222240448,
      "reward_std": 0.5695139020681381,
      "rewards/accuracy_reward": 0.5212053544819355,
      "rewards/format_reward": 0.9832589253783226,
      "rewards/log_scaled_reward": 0.2765395335154608,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 227.70425033569336,
      "epoch": 0.6119402985074627,
      "grad_norm": 0.8301727771759033,
      "learning_rate": 8.802029828000155e-07,
      "loss": 0.1365,
      "num_tokens": 22182102.0,
      "reward": 1.7540639638900757,
      "reward_std": 0.504084050655365,
      "rewards/accuracy_reward": 0.4899553582072258,
      "rewards/format_reward": 0.9754464253783226,
      "rewards/log_scaled_reward": 0.2886621206998825,
      "step": 41
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 186.12500953674316,
      "epoch": 0.6268656716417911,
      "grad_norm": 0.6249210238456726,
      "learning_rate": 8.71572412738697e-07,
      "loss": 0.1336,
      "num_tokens": 22471350.0,
      "reward": 1.9936908185482025,
      "reward_std": 0.6450418382883072,
      "rewards/accuracy_reward": 0.5926339402794838,
      "rewards/format_reward": 0.9888392761349678,
      "rewards/log_scaled_reward": 0.41221751645207405,
      "step": 42
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 169.09710693359375,
      "epoch": 0.6417910447761194,
      "grad_norm": 0.7566676139831543,
      "learning_rate": 8.626871855061437e-07,
      "loss": 0.1662,
      "num_tokens": 22758477.0,
      "reward": 1.8480805903673172,
      "reward_std": 0.5597276613116264,
      "rewards/accuracy_reward": 0.511160708963871,
      "rewards/format_reward": 0.9944196343421936,
      "rewards/log_scaled_reward": 0.3425001185387373,
      "step": 43
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 152.94978141784668,
      "epoch": 0.6567164179104478,
      "grad_norm": 0.9348928332328796,
      "learning_rate": 8.535533905932737e-07,
      "loss": 0.1287,
      "num_tokens": 23022432.0,
      "reward": 1.8545437455177307,
      "reward_std": 0.5482046529650688,
      "rewards/accuracy_reward": 0.5078125037252903,
      "rewards/format_reward": 0.9921874925494194,
      "rewards/log_scaled_reward": 0.35454366356134415,
      "step": 44
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 128.6517915725708,
      "epoch": 0.6716417910447762,
      "grad_norm": 1.1114394664764404,
      "learning_rate": 8.441772878468769e-07,
      "loss": 0.1287,
      "num_tokens": 23263912.0,
      "reward": 1.7271955758333206,
      "reward_std": 0.5110182501375675,
      "rewards/accuracy_reward": 0.4330357164144516,
      "rewards/format_reward": 0.9977678507566452,
      "rewards/log_scaled_reward": 0.29639193043112755,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 143.97545337677002,
      "epoch": 0.6865671641791045,
      "grad_norm": 1.1403673887252808,
      "learning_rate": 8.34565303179429e-07,
      "loss": 0.1742,
      "num_tokens": 23526706.0,
      "reward": 1.6235045939683914,
      "reward_std": 0.5148132182657719,
      "rewards/accuracy_reward": 0.386160708963871,
      "rewards/format_reward": 0.987723208963871,
      "rewards/log_scaled_reward": 0.24962060060352087,
      "step": 46
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 104.42187976837158,
      "epoch": 0.7014925373134329,
      "grad_norm": 1.6832914352416992,
      "learning_rate": 8.247240241650917e-07,
      "loss": 0.1296,
      "num_tokens": 23736324.0,
      "reward": 1.690138816833496,
      "reward_std": 0.5133109800517559,
      "rewards/accuracy_reward": 0.4051339328289032,
      "rewards/format_reward": 0.9933035671710968,
      "rewards/log_scaled_reward": 0.29170125164091587,
      "step": 47
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 92.65848731994629,
      "epoch": 0.7164179104477612,
      "grad_norm": 1.6501548290252686,
      "learning_rate": 8.146601955249187e-07,
      "loss": 0.1411,
      "num_tokens": 23946162.0,
      "reward": 1.6304273456335068,
      "reward_std": 0.49451132118701935,
      "rewards/accuracy_reward": 0.36941964365541935,
      "rewards/format_reward": 0.995535708963871,
      "rewards/log_scaled_reward": 0.26547193340957165,
      "step": 48
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 78.72098636627197,
      "epoch": 0.7313432835820896,
      "grad_norm": 2.571906566619873,
      "learning_rate": 8.043807145043603e-07,
      "loss": 0.1793,
      "num_tokens": 24153096.0,
      "reward": 1.581173524260521,
      "reward_std": 0.43424950167536736,
      "rewards/accuracy_reward": 0.33748282864689827,
      "rewards/format_reward": 0.9966517835855484,
      "rewards/log_scaled_reward": 0.24970022030174732,
      "step": 49
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 64.70759201049805,
      "epoch": 0.746268656716418,
      "grad_norm": 2.7052981853485107,
      "learning_rate": 7.938926261462365e-07,
      "loss": 0.1791,
      "num_tokens": 24349226.0,
      "reward": 1.5854334235191345,
      "reward_std": 0.3977060168981552,
      "rewards/accuracy_reward": 0.33593750186264515,
      "rewards/format_reward": 0.9955357015132904,
      "rewards/log_scaled_reward": 0.2539601270109415,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 46.40067148208618,
      "epoch": 0.7611940298507462,
      "grad_norm": 6.834611415863037,
      "learning_rate": 7.832031184624164e-07,
      "loss": 0.1055,
      "num_tokens": 24528585.0,
      "reward": 1.5368833392858505,
      "reward_std": 0.47740813344717026,
      "rewards/accuracy_reward": 0.30133928544819355,
      "rewards/format_reward": 0.9988839253783226,
      "rewards/log_scaled_reward": 0.2366600539535284,
      "step": 51
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 38.36384057998657,
      "epoch": 0.7761194029850746,
      "grad_norm": 4.969581127166748,
      "learning_rate": 7.723195175075135e-07,
      "loss": 0.1206,
      "num_tokens": 24691607.0,
      "reward": 1.415985830128193,
      "reward_std": 0.29450324457138777,
      "rewards/accuracy_reward": 0.2377232169965282,
      "rewards/format_reward": 0.9977678507566452,
      "rewards/log_scaled_reward": 0.18049467215314507,
      "step": 52
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 33.99218940734863,
      "epoch": 0.7910447761194029,
      "grad_norm": 4.195903778076172,
      "learning_rate": 7.612492823579744e-07,
      "loss": 0.145,
      "num_tokens": 24849736.0,
      "reward": 1.5035328567028046,
      "reward_std": 0.34667503647506237,
      "rewards/accuracy_reward": 0.27790178544819355,
      "rewards/format_reward": 0.9988839253783226,
      "rewards/log_scaled_reward": 0.2267470918595791,
      "step": 53
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 26.474331378936768,
      "epoch": 0.8059701492537313,
      "grad_norm": 24.98171615600586,
      "learning_rate": 7.5e-07,
      "loss": 0.0613,
      "num_tokens": 25008289.0,
      "reward": 1.5673803389072418,
      "reward_std": 0.28740744665265083,
      "rewards/accuracy_reward": 0.3069196417927742,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": 0.26046060863882303,
      "step": 54
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 25.195313692092896,
      "epoch": 0.8208955223880597,
      "grad_norm": 5.9815216064453125,
      "learning_rate": 7.385793801298042e-07,
      "loss": 0.0376,
      "num_tokens": 25169984.0,
      "reward": 1.4668240398168564,
      "reward_std": 0.3536365833133459,
      "rewards/accuracy_reward": 0.25369161926209927,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": 0.2157079027965665,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 23.91294765472412,
      "epoch": 0.835820895522388,
      "grad_norm": 7.0327372550964355,
      "learning_rate": 7.269952498697734e-07,
      "loss": 0.0386,
      "num_tokens": 25322466.0,
      "reward": 1.4718168079853058,
      "reward_std": 0.3211175389587879,
      "rewards/accuracy_reward": 0.2578124962747097,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": 0.21400425024330616,
      "step": 56
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 23.36049222946167,
      "epoch": 0.8507462686567164,
      "grad_norm": 5.700937271118164,
      "learning_rate": 7.152555484041475e-07,
      "loss": 0.0265,
      "num_tokens": 25466397.0,
      "reward": 1.5502240508794785,
      "reward_std": 0.2749287262558937,
      "rewards/accuracy_reward": 0.296875,
      "rewards/format_reward": 0.9988839253783226,
      "rewards/log_scaled_reward": 0.25446509197354317,
      "step": 57
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 23.949777364730835,
      "epoch": 0.8656716417910447,
      "grad_norm": 5.249198913574219,
      "learning_rate": 7.033683215379002e-07,
      "loss": 0.023,
      "num_tokens": 25613824.0,
      "reward": 1.5308541655540466,
      "reward_std": 0.26417338382452726,
      "rewards/accuracy_reward": 0.28794642724096775,
      "rewards/format_reward": 0.9988839253783226,
      "rewards/log_scaled_reward": 0.2440237421542406,
      "step": 58
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 23.375000953674316,
      "epoch": 0.8805970149253731,
      "grad_norm": 8.684850692749023,
      "learning_rate": 6.913417161825449e-07,
      "loss": 0.0289,
      "num_tokens": 25760176.0,
      "reward": 1.566646233201027,
      "reward_std": 0.22551130689680576,
      "rewards/accuracy_reward": 0.30691963620483875,
      "rewards/format_reward": 0.9966517761349678,
      "rewards/log_scaled_reward": 0.26307476311922073,
      "step": 59
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 21.989956378936768,
      "epoch": 0.8955223880597015,
      "grad_norm": 34.08607864379883,
      "learning_rate": 6.7918397477265e-07,
      "loss": 0.0144,
      "num_tokens": 25914911.0,
      "reward": 1.628076210618019,
      "reward_std": 0.3200679961591959,
      "rewards/accuracy_reward": 0.33705356903374195,
      "rewards/format_reward": 0.9966517761349678,
      "rewards/log_scaled_reward": 0.2943707942031324,
      "step": 60
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 22.574777841567993,
      "epoch": 0.9104477611940298,
      "grad_norm": 5.27999210357666,
      "learning_rate": 6.669034296168854e-07,
      "loss": 0.0217,
      "num_tokens": 26077314.0,
      "reward": 1.6822472661733627,
      "reward_std": 0.2845423389226198,
      "rewards/accuracy_reward": 0.3649553582072258,
      "rewards/format_reward": 0.995535708963871,
      "rewards/log_scaled_reward": 0.3217560853809118,
      "step": 61
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 22.198662042617798,
      "epoch": 0.9253731343283582,
      "grad_norm": 4.5566277503967285,
      "learning_rate": 6.545084971874736e-07,
      "loss": 0.0064,
      "num_tokens": 26222620.0,
      "reward": 1.5729791224002838,
      "reward_std": 0.22709419997408986,
      "rewards/accuracy_reward": 0.31138393096625805,
      "rewards/format_reward": 0.9921874925494194,
      "rewards/log_scaled_reward": 0.2694076579064131,
      "step": 62
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 21.32924246788025,
      "epoch": 0.9402985074626866,
      "grad_norm": 3.140645980834961,
      "learning_rate": 6.420076723519614e-07,
      "loss": 0.0068,
      "num_tokens": 26371243.0,
      "reward": 1.6745910048484802,
      "reward_std": 0.20502757839858532,
      "rewards/accuracy_reward": 0.3582589291036129,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": 0.3163319919258356,
      "step": 63
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 21.34709930419922,
      "epoch": 0.9552238805970149,
      "grad_norm": 4.071537494659424,
      "learning_rate": 6.294095225512604e-07,
      "loss": 0.0055,
      "num_tokens": 26527050.0,
      "reward": 1.5537814646959305,
      "reward_std": 0.30234235525131226,
      "rewards/accuracy_reward": 0.2979910681024194,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": 0.25579037982970476,
      "step": 64
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 21.24776864051819,
      "epoch": 0.9701492537313433,
      "grad_norm": 4.297541618347168,
      "learning_rate": 6.167226819279527e-07,
      "loss": 0.0114,
      "num_tokens": 26679280.0,
      "reward": 1.5483618080615997,
      "reward_std": 0.1530774086713791,
      "rewards/accuracy_reward": 0.2968749953433871,
      "rewards/format_reward": 0.9966517835855484,
      "rewards/log_scaled_reward": 0.25483495742082596,
      "step": 65
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 21.180555820465088,
      "epoch": 0.9850746268656716,
      "grad_norm": 3.6126675605773926,
      "learning_rate": 6.039558454088795e-07,
      "loss": 0.009,
      "num_tokens": 26828211.0,
      "reward": 1.6490006893873215,
      "reward_std": 0.29192496836185455,
      "rewards/accuracy_reward": 0.3459821417927742,
      "rewards/format_reward": 0.9988839253783226,
      "rewards/log_scaled_reward": 0.304134588688612,
      "step": 66
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 20.77455449104309,
      "epoch": 1.0149253731343284,
      "grad_norm": 4.067701816558838,
      "learning_rate": 5.911177627460738e-07,
      "loss": 0.0094,
      "num_tokens": 26965209.0,
      "reward": 1.7030873149633408,
      "reward_std": 0.2512226551771164,
      "rewards/accuracy_reward": 0.37388391979038715,
      "rewards/format_reward": 0.9966517761349678,
      "rewards/log_scaled_reward": 0.3325514607131481,
      "step": 67
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 20.35267925262451,
      "epoch": 1.0298507462686568,
      "grad_norm": 3.4311537742614746,
      "learning_rate": 5.782172325201155e-07,
      "loss": 0.0126,
      "num_tokens": 27116397.0,
      "reward": 1.5380910784006119,
      "reward_std": 0.20529233757406473,
      "rewards/accuracy_reward": 0.2901785708963871,
      "rewards/format_reward": 0.9988839253783226,
      "rewards/log_scaled_reward": 0.24902847222983837,
      "step": 68
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 20.400670528411865,
      "epoch": 1.044776119402985,
      "grad_norm": 4.965053081512451,
      "learning_rate": 5.652630961100258e-07,
      "loss": 0.0182,
      "num_tokens": 27260012.0,
      "reward": 1.4891109764575958,
      "reward_std": 0.2253081511007622,
      "rewards/accuracy_reward": 0.26562499813735485,
      "rewards/format_reward": 0.9988839253783226,
      "rewards/log_scaled_reward": 0.22460200637578964,
      "step": 69
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 19.54799175262451,
      "epoch": 1.0597014925373134,
      "grad_norm": 9.666987419128418,
      "learning_rate": 5.522642316338268e-07,
      "loss": 0.0249,
      "num_tokens": 27404823.0,
      "reward": 1.5430727303028107,
      "reward_std": 0.2664187829941511,
      "rewards/accuracy_reward": 0.2924107164144516,
      "rewards/format_reward": 0.9977678507566452,
      "rewards/log_scaled_reward": 0.25289412308484316,
      "step": 70
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 18.02120590209961,
      "epoch": 1.0746268656716418,
      "grad_norm": 15.713820457458496,
      "learning_rate": 5.392295478639225e-07,
      "loss": 0.0167,
      "num_tokens": 27555962.0,
      "reward": 1.1339266449213028,
      "reward_std": 0.34662946686148643,
      "rewards/accuracy_reward": 0.1026785708963871,
      "rewards/format_reward": 0.9676339328289032,
      "rewards/log_scaled_reward": 0.06361408122756984,
      "step": 71
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 19.39732265472412,
      "epoch": 1.0895522388059702,
      "grad_norm": 21.441844940185547,
      "learning_rate": 5.26167978121472e-07,
      "loss": 0.0653,
      "num_tokens": 27696134.0,
      "reward": 1.0387937128543854,
      "reward_std": 0.14305981155484915,
      "rewards/accuracy_reward": 0.04464285704307258,
      "rewards/format_reward": 0.9866071417927742,
      "rewards/log_scaled_reward": 0.0075436777296999935,
      "step": 72
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.95089340209961,
      "epoch": 1.1044776119402986,
      "grad_norm": 12.856225967407227,
      "learning_rate": 5.130884741539366e-07,
      "loss": 0.018,
      "num_tokens": 27843106.0,
      "reward": 0.9690398126840591,
      "reward_std": 0.01615892370318761,
      "rewards/accuracy_reward": 0.0022321429569274187,
      "rewards/format_reward": 0.9988839253783226,
      "rewards/log_scaled_reward": -0.03207633784040809,
      "step": 73
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.233259677886963,
      "epoch": 1.1194029850746268,
      "grad_norm": 1.9851570129394531,
      "learning_rate": 5e-07,
      "loss": 0.0015,
      "num_tokens": 27969987.0,
      "reward": 0.9649922177195549,
      "reward_std": 0.0031741062664423225,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.9988839253783226,
      "rewards/log_scaled_reward": -0.03389178216457367,
      "step": 74
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.01897418498993,
      "epoch": 1.1343283582089552,
      "grad_norm": 0.0,
      "learning_rate": 4.869115258460634e-07,
      "loss": 0.0,
      "num_tokens": 28107828.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 75
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 14.996652722358704,
      "epoch": 1.1492537313432836,
      "grad_norm": 1.0442914962768555,
      "learning_rate": 4.7383202187852804e-07,
      "loss": -0.0006,
      "num_tokens": 28248081.0,
      "reward": 0.9638938158750534,
      "reward_std": 0.006288029253482819,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.9977678507566452,
      "rewards/log_scaled_reward": -0.03387411683797836,
      "step": 76
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.00334918498993,
      "epoch": 1.164179104477612,
      "grad_norm": 1.5838154554367065,
      "learning_rate": 4.6077045213607755e-07,
      "loss": 0.0005,
      "num_tokens": 28386956.0,
      "reward": 0.9627697318792343,
      "reward_std": 0.009467384777963161,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.9966517835855484,
      "rewards/log_scaled_reward": -0.0338821173645556,
      "step": 77
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.1791044776119404,
      "grad_norm": 1.337672472000122,
      "learning_rate": 4.477357683661733e-07,
      "loss": -0.0,
      "num_tokens": 28527204.0,
      "reward": 0.9638830795884132,
      "reward_std": 0.006318369880318642,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.9977678582072258,
      "rewards/log_scaled_reward": -0.033884843811392784,
      "step": 78
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.1940298507462686,
      "grad_norm": 0.0,
      "learning_rate": 4.347369038899743e-07,
      "loss": 0.0,
      "num_tokens": 28671788.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 79
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.208955223880597,
      "grad_norm": 0.0,
      "learning_rate": 4.2178276747988444e-07,
      "loss": 0.0,
      "num_tokens": 28811644.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 80
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.2238805970149254,
      "grad_norm": 0.0,
      "learning_rate": 4.0888223725392624e-07,
      "loss": 0.0,
      "num_tokens": 28949996.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 81
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.2388059701492538,
      "grad_norm": 0.0,
      "learning_rate": 3.960441545911204e-07,
      "loss": 0.0,
      "num_tokens": 29100540.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 82
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.2537313432835822,
      "grad_norm": 0.9363570213317871,
      "learning_rate": 3.8327731807204744e-07,
      "loss": -0.0,
      "num_tokens": 29242748.0,
      "reward": 0.9650017619132996,
      "reward_std": 0.0031542566139250994,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.9988839253783226,
      "rewards/log_scaled_reward": -0.03388223238289356,
      "step": 83
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.2686567164179103,
      "grad_norm": 0.0,
      "learning_rate": 3.7059047744873955e-07,
      "loss": 0.0,
      "num_tokens": 29394812.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 84
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.2835820895522387,
      "grad_norm": 0.0,
      "learning_rate": 3.5799232764803867e-07,
      "loss": 0.0,
      "num_tokens": 29526756.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 85
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.2985074626865671,
      "grad_norm": 0.0,
      "learning_rate": 3.454915028125263e-07,
      "loss": 0.0,
      "num_tokens": 29662892.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 86
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.3134328358208955,
      "grad_norm": 0.0,
      "learning_rate": 3.330965703831146e-07,
      "loss": 0.0,
      "num_tokens": 29808492.0,
      "reward": 0.975348062813282,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.024652006570249796,
      "step": 87
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.328358208955224,
      "grad_norm": 0.0,
      "learning_rate": 3.2081602522734985e-07,
      "loss": 0.0,
      "num_tokens": 29960316.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 88
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.3432835820895521,
      "grad_norm": 0.0,
      "learning_rate": 3.086582838174551e-07,
      "loss": 0.0,
      "num_tokens": 30099644.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 89
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.005581259727478,
      "epoch": 1.3582089552238805,
      "grad_norm": 0.0,
      "learning_rate": 2.9663167846209996e-07,
      "loss": 0.0,
      "num_tokens": 30246217.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 90
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.373134328358209,
      "grad_norm": 0.0,
      "learning_rate": 2.847444515958523e-07,
      "loss": 0.0,
      "num_tokens": 30391785.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 91
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.3880597014925373,
      "grad_norm": 0.0,
      "learning_rate": 2.730047501302266e-07,
      "loss": 0.0,
      "num_tokens": 30531289.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 92
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.4029850746268657,
      "grad_norm": 3.6358273029327393,
      "learning_rate": 2.6142061987019574e-07,
      "loss": -0.0,
      "num_tokens": 30664633.0,
      "reward": 0.966103158891201,
      "reward_std": 1.4753467439732049e-05,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.033896906301379204,
      "step": 93
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.417910447761194,
      "grad_norm": 0.0,
      "learning_rate": 2.500000000000001e-07,
      "loss": 0.0,
      "num_tokens": 30820657.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 94
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.4328358208955223,
      "grad_norm": 0.0,
      "learning_rate": 2.387507176420256e-07,
      "loss": 0.0,
      "num_tokens": 30965737.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 95
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.00111699104309,
      "epoch": 1.4477611940298507,
      "grad_norm": 32.66771697998047,
      "learning_rate": 2.2768048249248644e-07,
      "loss": -0.0002,
      "num_tokens": 31108402.0,
      "reward": 0.9649775922298431,
      "reward_std": 0.0031557143665850163,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 0.9988839253783226,
      "rewards/log_scaled_reward": -0.03390640066936612,
      "step": 96
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.462686567164179,
      "grad_norm": 0.0,
      "learning_rate": 2.167968815375837e-07,
      "loss": 0.0,
      "num_tokens": 31257634.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 97
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.4776119402985075,
      "grad_norm": 0.0,
      "learning_rate": 2.0610737385376348e-07,
      "loss": 0.0,
      "num_tokens": 31412970.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 98
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.4925373134328357,
      "grad_norm": 0.0,
      "learning_rate": 1.9561928549563966e-07,
      "loss": 0.0,
      "num_tokens": 31564498.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 99
    },
    {
      "epoch": 1.5074626865671643,
      "grad_norm": 0.0,
      "learning_rate": 1.8533980447508135e-07,
      "loss": 0.0,
      "step": 100
    },
    {
      "epoch": 1.5074626865671643,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 15.000150587305676,
      "eval_loss": 0.0,
      "eval_num_tokens": 31697802.0,
      "eval_reward": 0.9661169648170471,
      "eval_reward_std": 0.0,
      "eval_rewards/accuracy_reward": 0.0,
      "eval_rewards/format_reward": 1.0,
      "eval_rewards/log_scaled_reward": -0.03388310596346855,
      "eval_runtime": 724.0402,
      "eval_samples_per_second": 6.906,
      "eval_steps_per_second": 0.062,
      "step": 100
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.5223880597014925,
      "grad_norm": 0.0,
      "learning_rate": 1.7527597583490823e-07,
      "loss": 0.0,
      "num_tokens": 31837018.0,
      "reward": 0.9707325138151646,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.029267556266859174,
      "step": 101
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.537313432835821,
      "grad_norm": 0.0,
      "learning_rate": 1.6543469682057104e-07,
      "loss": 0.0,
      "num_tokens": 31986642.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 102
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.5522388059701493,
      "grad_norm": 0.0,
      "learning_rate": 1.5582271215312293e-07,
      "loss": 0.0,
      "num_tokens": 32126642.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 103
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.5671641791044775,
      "grad_norm": 0.0,
      "learning_rate": 1.4644660940672627e-07,
      "loss": 0.0,
      "num_tokens": 32276154.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 104
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.582089552238806,
      "grad_norm": 0.0,
      "learning_rate": 1.3731281449385628e-07,
      "loss": 0.0,
      "num_tokens": 32426898.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 105
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.5970149253731343,
      "grad_norm": 0.0,
      "learning_rate": 1.284275872613028e-07,
      "loss": 0.0,
      "num_tokens": 32571226.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 106
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.6119402985074627,
      "grad_norm": 0.0,
      "learning_rate": 1.1979701719998454e-07,
      "loss": 0.0,
      "num_tokens": 32719426.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 107
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.626865671641791,
      "grad_norm": 0.0,
      "learning_rate": 1.1142701927151454e-07,
      "loss": 0.0,
      "num_tokens": 32852794.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 108
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.6417910447761193,
      "grad_norm": 0.0,
      "learning_rate": 1.0332332985438247e-07,
      "loss": 0.0,
      "num_tokens": 33000498.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 109
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.6567164179104479,
      "grad_norm": 0.0,
      "learning_rate": 9.549150281252632e-08,
      "loss": 0.0,
      "num_tokens": 33139850.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 110
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.671641791044776,
      "grad_norm": 0.0,
      "learning_rate": 8.793690568899215e-08,
      "loss": 0.0,
      "num_tokens": 33276866.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 111
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.6865671641791045,
      "grad_norm": 0.0,
      "learning_rate": 8.066471602728803e-08,
      "loss": 0.0,
      "num_tokens": 33407970.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 112
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.7014925373134329,
      "grad_norm": 0.0,
      "learning_rate": 7.36799178229539e-08,
      "loss": 0.0,
      "num_tokens": 33549218.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 113
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.716417910447761,
      "grad_norm": 0.0,
      "learning_rate": 6.698729810778064e-08,
      "loss": 0.0,
      "num_tokens": 33694498.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 114
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.7313432835820897,
      "grad_norm": 0.0,
      "learning_rate": 6.059144366901736e-08,
      "loss": 0.0,
      "num_tokens": 33836482.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 115
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.7462686567164178,
      "grad_norm": 0.0,
      "learning_rate": 5.44967379058161e-08,
      "loss": 0.0,
      "num_tokens": 33970506.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 116
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.002233028411865,
      "epoch": 1.7611940298507462,
      "grad_norm": 0.0,
      "learning_rate": 4.870735782506979e-08,
      "loss": 0.0,
      "num_tokens": 34139396.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 117
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.7761194029850746,
      "grad_norm": 0.0,
      "learning_rate": 4.322727117869951e-08,
      "loss": 0.0,
      "num_tokens": 34279244.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 118
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.7910447761194028,
      "grad_norm": 0.0,
      "learning_rate": 3.806023374435663e-08,
      "loss": 0.0,
      "num_tokens": 34418092.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 119
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.8059701492537314,
      "grad_norm": 0.0,
      "learning_rate": 3.3209786751399184e-08,
      "loss": 0.0,
      "num_tokens": 34570476.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 120
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.8208955223880596,
      "grad_norm": 0.0,
      "learning_rate": 2.8679254453910785e-08,
      "loss": 0.0,
      "num_tokens": 34714220.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 121
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.835820895522388,
      "grad_norm": 0.0,
      "learning_rate": 2.4471741852423233e-08,
      "loss": 0.0,
      "num_tokens": 34874452.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 122
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.8507462686567164,
      "grad_norm": 0.0,
      "learning_rate": 2.0590132565903473e-08,
      "loss": 0.0,
      "num_tokens": 35020932.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 123
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.8656716417910446,
      "grad_norm": 0.0,
      "learning_rate": 1.7037086855465898e-08,
      "loss": 0.0,
      "num_tokens": 35158836.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 124
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.8805970149253732,
      "grad_norm": 0.0,
      "learning_rate": 1.3815039801161722e-08,
      "loss": 0.0,
      "num_tokens": 35299044.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 125
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.8955223880597014,
      "grad_norm": 0.0,
      "learning_rate": 1.0926199633097154e-08,
      "loss": 0.0,
      "num_tokens": 35438660.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 126
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.9104477611940298,
      "grad_norm": 0.0,
      "learning_rate": 8.372546218022746e-09,
      "loss": 0.0,
      "num_tokens": 35589036.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 127
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.9253731343283582,
      "grad_norm": 0.0,
      "learning_rate": 6.15582970243117e-09,
      "loss": 0.0,
      "num_tokens": 35732844.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 128
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.9402985074626866,
      "grad_norm": 0.0,
      "learning_rate": 4.277569313094809e-09,
      "loss": 0.0,
      "num_tokens": 35869612.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 129
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.955223880597015,
      "grad_norm": 0.0,
      "learning_rate": 2.739052315863355e-09,
      "loss": 0.0,
      "num_tokens": 36024796.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 130
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.000000953674316,
      "epoch": 1.9701492537313432,
      "grad_norm": 0.0,
      "learning_rate": 1.541333133436018e-09,
      "loss": 0.0,
      "num_tokens": 36165404.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 131
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 15.0,
      "epoch": 1.9850746268656716,
      "grad_norm": 0.0,
      "learning_rate": 6.852326227130833e-10,
      "loss": 0.0,
      "num_tokens": 36316348.0,
      "reward": 0.9661169648170471,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.0,
      "rewards/format_reward": 1.0,
      "rewards/log_scaled_reward": -0.03388310596346855,
      "step": 132
    },
    {
      "epoch": 1.9850746268656716,
      "step": 132,
      "total_flos": 0.0,
      "train_loss": 0.04579740530018937,
      "train_runtime": 16666.94,
      "train_samples_per_second": 0.9,
      "train_steps_per_second": 0.008
    }
  ],
  "logging_steps": 1,
  "max_steps": 134,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}