{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9936628643852978,
  "eval_steps": 100,
  "global_step": 196,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 544.7745819091797,
      "epoch": 0.010139416983523447,
      "grad_norm": 0.2261611670255661,
      "learning_rate": 5e-08,
      "loss": -0.2592,
      "num_tokens": 636118.0,
      "reward": -0.6724561750888824,
      "reward_std": 0.4402390718460083,
      "rewards/cosine_scaled_reward": -0.6758043766021729,
      "rewards/format_reward": 0.003348214435391128,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 549.2812843322754,
      "epoch": 0.020278833967046894,
      "grad_norm": 0.22706830501556396,
      "learning_rate": 1e-07,
      "loss": -0.2276,
      "num_tokens": 1263330.0,
      "reward": -0.6239746958017349,
      "reward_std": 0.49842649325728416,
      "rewards/cosine_scaled_reward": -0.6317871585488319,
      "rewards/format_reward": 0.007812500349245965,
      "step": 2
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 569.3448944091797,
      "epoch": 0.030418250950570342,
      "grad_norm": 0.18996624648571014,
      "learning_rate": 1.5e-07,
      "loss": -0.2165,
      "num_tokens": 1913583.0,
      "reward": -0.614225223660469,
      "reward_std": 0.4692864343523979,
      "rewards/cosine_scaled_reward": -0.6231537759304047,
      "rewards/format_reward": 0.008928571827709675,
      "step": 3
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 539.3951110839844,
      "epoch": 0.04055766793409379,
      "grad_norm": 0.22352048754692078,
      "learning_rate": 2e-07,
      "loss": -0.2585,
      "num_tokens": 2533833.0,
      "reward": -0.6429695598781109,
      "reward_std": 0.4445586446672678,
      "rewards/cosine_scaled_reward": -0.6507820263504982,
      "rewards/format_reward": 0.007812500349245965,
      "step": 4
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 576.1462287902832,
      "epoch": 0.050697084917617236,
      "grad_norm": 0.21681766211986542,
      "learning_rate": 2.5e-07,
      "loss": -0.2409,
      "num_tokens": 3176844.0,
      "reward": -0.6119992136955261,
      "reward_std": 0.4665570706129074,
      "rewards/cosine_scaled_reward": -0.6164634898304939,
      "rewards/format_reward": 0.004464285913854837,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 570.2154197692871,
      "epoch": 0.060836501901140684,
      "grad_norm": 0.21843321621418,
      "learning_rate": 3e-07,
      "loss": -0.2408,
      "num_tokens": 3816949.0,
      "reward": -0.5715240500867367,
      "reward_std": 0.5433307699859142,
      "rewards/cosine_scaled_reward": -0.5771043859422207,
      "rewards/format_reward": 0.005580357392318547,
      "step": 6
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 539.3672180175781,
      "epoch": 0.07097591888466413,
      "grad_norm": 0.21752215921878815,
      "learning_rate": 3.5e-07,
      "loss": -0.1852,
      "num_tokens": 4424646.0,
      "reward": -0.5407312288880348,
      "reward_std": 0.5957661867141724,
      "rewards/cosine_scaled_reward": -0.5518919117748737,
      "rewards/format_reward": 0.011160714668221772,
      "step": 7
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 558.9989128112793,
      "epoch": 0.08111533586818757,
      "grad_norm": 0.21230322122573853,
      "learning_rate": 4e-07,
      "loss": -0.205,
      "num_tokens": 5053061.0,
      "reward": -0.6614980250597,
      "reward_std": 0.47641005739569664,
      "rewards/cosine_scaled_reward": -0.667078360915184,
      "rewards/format_reward": 0.005580357392318547,
      "step": 8
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 576.405158996582,
      "epoch": 0.09125475285171103,
      "grad_norm": 0.22808963060379028,
      "learning_rate": 4.5e-07,
      "loss": -0.2588,
      "num_tokens": 5692848.0,
      "reward": -0.6227121576666832,
      "reward_std": 0.45661213248968124,
      "rewards/cosine_scaled_reward": -0.6294085532426834,
      "rewards/format_reward": 0.006696428870782256,
      "step": 9
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 516.9732475280762,
      "epoch": 0.10139416983523447,
      "grad_norm": 0.2245020717382431,
      "learning_rate": 5e-07,
      "loss": -0.1928,
      "num_tokens": 6282784.0,
      "reward": -0.5244979597628117,
      "reward_std": 0.5999544225633144,
      "rewards/cosine_scaled_reward": -0.5278461389243603,
      "rewards/format_reward": 0.003348214435391128,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.3047180175781,
      "epoch": 0.11153358681875793,
      "grad_norm": 0.22737731039524078,
      "learning_rate": 5.5e-07,
      "loss": -0.2092,
      "num_tokens": 6942961.0,
      "reward": -0.609170563519001,
      "reward_std": 0.47449312917888165,
      "rewards/cosine_scaled_reward": -0.6125187538564205,
      "rewards/format_reward": 0.003348214435391128,
      "step": 11
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 587.389533996582,
      "epoch": 0.12167300380228137,
      "grad_norm": 0.22877997159957886,
      "learning_rate": 6e-07,
      "loss": -0.2051,
      "num_tokens": 7600574.0,
      "reward": -0.5761966481804848,
      "reward_std": 0.47745756432414055,
      "rewards/cosine_scaled_reward": -0.5884733907878399,
      "rewards/format_reward": 0.012276786146685481,
      "step": 12
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 605.9844055175781,
      "epoch": 0.13181242078580482,
      "grad_norm": 0.20151640474796295,
      "learning_rate": 6.5e-07,
      "loss": -0.1612,
      "num_tokens": 8271720.0,
      "reward": -0.5711147747933865,
      "reward_std": 0.5248831808567047,
      "rewards/cosine_scaled_reward": -0.5789272524416447,
      "rewards/format_reward": 0.007812500349245965,
      "step": 13
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 554.7109680175781,
      "epoch": 0.14195183776932827,
      "grad_norm": 0.28496649861335754,
      "learning_rate": 7e-07,
      "loss": -0.1814,
      "num_tokens": 8900917.0,
      "reward": -0.5377125255763531,
      "reward_std": 0.589600820094347,
      "rewards/cosine_scaled_reward": -0.5481268018484116,
      "rewards/format_reward": 0.004464285913854837,
      "step": 14
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 597.1194534301758,
      "epoch": 0.1520912547528517,
      "grad_norm": 0.20674261450767517,
      "learning_rate": 7.5e-07,
      "loss": -0.1812,
      "num_tokens": 9564376.0,
      "reward": -0.5252576433122158,
      "reward_std": 0.5734145380556583,
      "rewards/cosine_scaled_reward": -0.5308379791676998,
      "rewards/format_reward": 0.005580357392318547,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 587.7846221923828,
      "epoch": 0.16223067173637515,
      "grad_norm": 0.18769149482250214,
      "learning_rate": 8e-07,
      "loss": -0.1947,
      "num_tokens": 10217959.0,
      "reward": -0.5038141719996929,
      "reward_std": 0.6175409816205502,
      "rewards/cosine_scaled_reward": -0.5071623660624027,
      "rewards/format_reward": 0.003348214435391128,
      "step": 16
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 625.3683319091797,
      "epoch": 0.17237008871989862,
      "grad_norm": 0.3610371947288513,
      "learning_rate": 8.499999999999999e-07,
      "loss": -0.1733,
      "num_tokens": 10912801.0,
      "reward": -0.5226632095873356,
      "reward_std": 0.5681972987949848,
      "rewards/cosine_scaled_reward": -0.5282435454428196,
      "rewards/format_reward": 0.005580357392318547,
      "step": 17
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 632.5993499755859,
      "epoch": 0.18250950570342206,
      "grad_norm": 0.183369979262352,
      "learning_rate": 9e-07,
      "loss": -0.1304,
      "num_tokens": 11602954.0,
      "reward": -0.4399515353143215,
      "reward_std": 0.6363755837082863,
      "rewards/cosine_scaled_reward": -0.44553186371922493,
      "rewards/format_reward": 0.005580357392318547,
      "step": 18
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 607.8593978881836,
      "epoch": 0.1926489226869455,
      "grad_norm": 0.18773187696933746,
      "learning_rate": 9.499999999999999e-07,
      "loss": -0.1421,
      "num_tokens": 12279580.0,
      "reward": -0.31881226040422916,
      "reward_std": 0.6840411424636841,
      "rewards/cosine_scaled_reward": -0.32439261162653565,
      "rewards/format_reward": 0.005580357392318547,
      "step": 19
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 647.7310485839844,
      "epoch": 0.20278833967046894,
      "grad_norm": 0.1795511543750763,
      "learning_rate": 1e-06,
      "loss": -0.0997,
      "num_tokens": 12980371.0,
      "reward": -0.29815168026834726,
      "reward_std": 0.6471684239804745,
      "rewards/cosine_scaled_reward": -0.30038381088525057,
      "rewards/format_reward": 0.0022321429569274187,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 660.1004867553711,
      "epoch": 0.21292775665399238,
      "grad_norm": 0.1688115894794464,
      "learning_rate": 9.999203468625015e-07,
      "loss": -0.0604,
      "num_tokens": 13693093.0,
      "reward": -0.11793683946598321,
      "reward_std": 0.6547496430575848,
      "rewards/cosine_scaled_reward": -0.11905290477443486,
      "rewards/format_reward": 0.0011160714784637094,
      "step": 21
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 716.8471298217773,
      "epoch": 0.22306717363751585,
      "grad_norm": 0.1519731730222702,
      "learning_rate": 9.99681412828496e-07,
      "loss": -0.0818,
      "num_tokens": 14468548.0,
      "reward": -0.2080770255997777,
      "reward_std": 0.5903773456811905,
      "rewards/cosine_scaled_reward": -0.21030915481969714,
      "rewards/format_reward": 0.0022321429569274187,
      "step": 22
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 675.1094131469727,
      "epoch": 0.2332065906210393,
      "grad_norm": 0.16422481834888458,
      "learning_rate": 9.992832740253644e-07,
      "loss": -0.0618,
      "num_tokens": 15211694.0,
      "reward": -0.13133149535860866,
      "reward_std": 0.6078522019088268,
      "rewards/cosine_scaled_reward": -0.1324475615983829,
      "rewards/format_reward": 0.0011160714784637094,
      "step": 23
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 732.0714569091797,
      "epoch": 0.24334600760456274,
      "grad_norm": 0.14371691644191742,
      "learning_rate": 9.987260573051267e-07,
      "loss": -0.0698,
      "num_tokens": 15997174.0,
      "reward": -0.12170885177329183,
      "reward_std": 0.5266584008932114,
      "rewards/cosine_scaled_reward": -0.12282492127269506,
      "rewards/format_reward": 0.0011160714784637094,
      "step": 24
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 674.2221298217773,
      "epoch": 0.2534854245880862,
      "grad_norm": 0.4856448471546173,
      "learning_rate": 9.98009940204023e-07,
      "loss": -0.047,
      "num_tokens": 16732605.0,
      "reward": 0.00843854621052742,
      "reward_std": 0.524466548115015,
      "rewards/cosine_scaled_reward": 0.00843854807317257,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 704.0792770385742,
      "epoch": 0.26362484157160965,
      "grad_norm": 0.15166230499744415,
      "learning_rate": 9.971351508859486e-07,
      "loss": -0.0499,
      "num_tokens": 17482292.0,
      "reward": 0.026816920144483447,
      "reward_std": 0.5594276264309883,
      "rewards/cosine_scaled_reward": 0.02570084948092699,
      "rewards/format_reward": 0.0011160714784637094,
      "step": 26
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 676.5524826049805,
      "epoch": 0.2737642585551331,
      "grad_norm": 0.2196025401353836,
      "learning_rate": 9.961019680697591e-07,
      "loss": -0.0143,
      "num_tokens": 18216659.0,
      "reward": -0.00026240432634949684,
      "reward_std": 0.587770089507103,
      "rewards/cosine_scaled_reward": -0.0002624071203172207,
      "rewards/format_reward": 0.0,
      "step": 27
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 751.2332916259766,
      "epoch": 0.28390367553865653,
      "grad_norm": 1.717673420906067,
      "learning_rate": 9.949107209404663e-07,
      "loss": 0.0007,
      "num_tokens": 19023404.0,
      "reward": 0.06609318475238979,
      "reward_std": 0.5609292760491371,
      "rewards/cosine_scaled_reward": 0.06609318096889183,
      "rewards/format_reward": 0.0,
      "step": 28
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 717.3995819091797,
      "epoch": 0.29404309252217997,
      "grad_norm": 0.1441808044910431,
      "learning_rate": 9.935617890443554e-07,
      "loss": -0.0205,
      "num_tokens": 19806930.0,
      "reward": 0.054041724652051926,
      "reward_std": 0.5101928748190403,
      "rewards/cosine_scaled_reward": 0.05404172372072935,
      "rewards/format_reward": 0.0,
      "step": 29
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 805.4955749511719,
      "epoch": 0.3041825095057034,
      "grad_norm": 0.12774203717708588,
      "learning_rate": 9.92055602168058e-07,
      "loss": -0.007,
      "num_tokens": 20674142.0,
      "reward": 0.047311737202107906,
      "reward_std": 0.4787377752363682,
      "rewards/cosine_scaled_reward": 0.04731173254549503,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 771.9632110595703,
      "epoch": 0.31432192648922685,
      "grad_norm": 0.1265403777360916,
      "learning_rate": 9.90392640201615e-07,
      "loss": 0.0387,
      "num_tokens": 21493429.0,
      "reward": 0.13890161426388659,
      "reward_std": 0.52209198102355,
      "rewards/cosine_scaled_reward": 0.1389016114990227,
      "rewards/format_reward": 0.0,
      "step": 31
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 763.1551666259766,
      "epoch": 0.3244613434727503,
      "grad_norm": 0.7253857851028442,
      "learning_rate": 9.885734329855797e-07,
      "loss": 0.023,
      "num_tokens": 22296960.0,
      "reward": 0.20044966787099838,
      "reward_std": 0.4556784965097904,
      "rewards/cosine_scaled_reward": 0.20044966042041779,
      "rewards/format_reward": 0.0,
      "step": 32
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 778.8873138427734,
      "epoch": 0.33460076045627374,
      "grad_norm": 0.1363135427236557,
      "learning_rate": 9.865985601422017e-07,
      "loss": -0.0147,
      "num_tokens": 23128899.0,
      "reward": 0.05959217494819313,
      "reward_std": 0.4793633334338665,
      "rewards/cosine_scaled_reward": 0.059291763755027205,
      "rewards/format_reward": 0.0,
      "step": 33
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 773.6607513427734,
      "epoch": 0.34474017743979724,
      "grad_norm": 0.12697336077690125,
      "learning_rate": 9.844686508907537e-07,
      "loss": 0.0046,
      "num_tokens": 23949587.0,
      "reward": 0.15802541188895702,
      "reward_std": 0.4121544286608696,
      "rewards/cosine_scaled_reward": 0.15802540606819093,
      "rewards/format_reward": 0.0,
      "step": 34
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 755.8393173217773,
      "epoch": 0.3548795944233207,
      "grad_norm": 0.13817483186721802,
      "learning_rate": 9.821843838470534e-07,
      "loss": 0.002,
      "num_tokens": 24769547.0,
      "reward": 0.15929587977007031,
      "reward_std": 0.4162367284297943,
      "rewards/cosine_scaled_reward": 0.159295872785151,
      "rewards/format_reward": 0.0,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 758.3839645385742,
      "epoch": 0.3650190114068441,
      "grad_norm": 0.1282864660024643,
      "learning_rate": 9.797464868072486e-07,
      "loss": 0.0237,
      "num_tokens": 25577171.0,
      "reward": 0.193751014303416,
      "reward_std": 0.48775994032621384,
      "rewards/cosine_scaled_reward": 0.193751011043787,
      "rewards/format_reward": 0.0,
      "step": 36
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 737.6127624511719,
      "epoch": 0.37515842839036756,
      "grad_norm": 0.13303731381893158,
      "learning_rate": 9.771557365159319e-07,
      "loss": 0.0083,
      "num_tokens": 26363600.0,
      "reward": 0.13768478552810848,
      "reward_std": 0.40786123648285866,
      "rewards/cosine_scaled_reward": 0.1376847765641287,
      "rewards/format_reward": 0.0,
      "step": 37
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 753.794677734375,
      "epoch": 0.385297845373891,
      "grad_norm": 0.137796089053154,
      "learning_rate": 9.744129584186597e-07,
      "loss": 0.0127,
      "num_tokens": 27165608.0,
      "reward": 0.12427436339203268,
      "reward_std": 0.47414594888687134,
      "rewards/cosine_scaled_reward": 0.12427435873541981,
      "rewards/format_reward": 0.0,
      "step": 38
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 799.005615234375,
      "epoch": 0.39543726235741444,
      "grad_norm": 0.1327836662530899,
      "learning_rate": 9.71519026398956e-07,
      "loss": 0.0341,
      "num_tokens": 28019301.0,
      "reward": 0.12463877908885479,
      "reward_std": 0.37975190207362175,
      "rewards/cosine_scaled_reward": 0.12463877722620964,
      "rewards/format_reward": 0.0,
      "step": 39
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 732.2611846923828,
      "epoch": 0.4055766793409379,
      "grad_norm": 0.14108538627624512,
      "learning_rate": 9.68474862499881e-07,
      "loss": 0.0519,
      "num_tokens": 28794671.0,
      "reward": 0.17419127095490694,
      "reward_std": 0.44464627653360367,
      "rewards/cosine_scaled_reward": 0.17419126071035862,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 725.4620819091797,
      "epoch": 0.4157160963244613,
      "grad_norm": 0.13821274042129517,
      "learning_rate": 9.652814366302568e-07,
      "loss": 0.0465,
      "num_tokens": 29570661.0,
      "reward": 0.32581639010459185,
      "reward_std": 0.3625021446496248,
      "rewards/cosine_scaled_reward": 0.325816385447979,
      "rewards/format_reward": 0.0,
      "step": 41
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 730.3828430175781,
      "epoch": 0.42585551330798477,
      "grad_norm": 0.1342957466840744,
      "learning_rate": 9.619397662556433e-07,
      "loss": 0.0505,
      "num_tokens": 30354052.0,
      "reward": 0.23176710167899728,
      "reward_std": 0.36819201707839966,
      "rewards/cosine_scaled_reward": 0.23176709050312638,
      "rewards/format_reward": 0.0,
      "step": 42
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 795.7176742553711,
      "epoch": 0.43599493029150826,
      "grad_norm": 0.12134254723787308,
      "learning_rate": 9.5845091607416e-07,
      "loss": 0.041,
      "num_tokens": 31185719.0,
      "reward": 0.2366194719215855,
      "reward_std": 0.4014102444052696,
      "rewards/cosine_scaled_reward": 0.23661946202628314,
      "rewards/format_reward": 0.0,
      "step": 43
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 801.1931076049805,
      "epoch": 0.4461343472750317,
      "grad_norm": 0.7807191014289856,
      "learning_rate": 9.548159976772592e-07,
      "loss": 0.0134,
      "num_tokens": 32033636.0,
      "reward": 0.21036180562805384,
      "reward_std": 0.40362199395895004,
      "rewards/cosine_scaled_reward": 0.21036179037764668,
      "rewards/format_reward": 0.0,
      "step": 44
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 819.1339569091797,
      "epoch": 0.45627376425855515,
      "grad_norm": 0.12294432520866394,
      "learning_rate": 9.510361691955606e-07,
      "loss": 0.021,
      "num_tokens": 32890028.0,
      "reward": 0.12139649732853286,
      "reward_std": 0.4755956120789051,
      "rewards/cosine_scaled_reward": 0.12139649360324256,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 717.7433319091797,
      "epoch": 0.4664131812420786,
      "grad_norm": 0.3100641071796417,
      "learning_rate": 9.471126349298556e-07,
      "loss": 0.0361,
      "num_tokens": 33663342.0,
      "reward": 0.23560622334480286,
      "reward_std": 0.42723193392157555,
      "rewards/cosine_scaled_reward": 0.23560621310025454,
      "rewards/format_reward": 0.0,
      "step": 46
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 762.2913284301758,
      "epoch": 0.47655259822560203,
      "grad_norm": 0.12677805125713348,
      "learning_rate": 9.430466449674013e-07,
      "loss": -0.0045,
      "num_tokens": 34494707.0,
      "reward": 0.1484816954471171,
      "reward_std": 0.3901702091097832,
      "rewards/cosine_scaled_reward": 0.14848168776370585,
      "rewards/format_reward": 0.0,
      "step": 47
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 789.130615234375,
      "epoch": 0.4866920152091255,
      "grad_norm": 0.12980540096759796,
      "learning_rate": 9.388394947836278e-07,
      "loss": -0.008,
      "num_tokens": 35328504.0,
      "reward": 0.09070475457701832,
      "reward_std": 0.3948558494448662,
      "rewards/cosine_scaled_reward": 0.09070474898908287,
      "rewards/format_reward": 0.0,
      "step": 48
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 792.5145416259766,
      "epoch": 0.4968314321926489,
      "grad_norm": 0.1316457837820053,
      "learning_rate": 9.344925248293835e-07,
      "loss": 0.0133,
      "num_tokens": 36159605.0,
      "reward": 0.2209687788854353,
      "reward_std": 0.39467399939894676,
      "rewards/cosine_scaled_reward": 0.22096876305295154,
      "rewards/format_reward": 0.0,
      "step": 49
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 776.4721374511719,
      "epoch": 0.5069708491761724,
      "grad_norm": 0.12961426377296448,
      "learning_rate": 9.300071201038501e-07,
      "loss": -0.0046,
      "num_tokens": 36981500.0,
      "reward": 0.15061184321530163,
      "reward_std": 0.4538601525127888,
      "rewards/cosine_scaled_reward": 0.15061183541547507,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 665.3649826049805,
      "epoch": 0.5171102661596958,
      "grad_norm": 0.1373734474182129,
      "learning_rate": 9.253847097132655e-07,
      "loss": 0.0202,
      "num_tokens": 37708491.0,
      "reward": 0.25066345604136586,
      "reward_std": 0.41370486095547676,
      "rewards/cosine_scaled_reward": 0.25066344952210784,
      "rewards/format_reward": 0.0,
      "step": 51
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 721.5212249755859,
      "epoch": 0.5272496831432193,
      "grad_norm": 0.12883824110031128,
      "learning_rate": 9.206267664155906e-07,
      "loss": 0.0295,
      "num_tokens": 38475350.0,
      "reward": 0.2094368627294898,
      "reward_std": 0.4836310148239136,
      "rewards/cosine_scaled_reward": 0.2094368515536189,
      "rewards/format_reward": 0.0,
      "step": 52
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 748.3538284301758,
      "epoch": 0.5373891001267427,
      "grad_norm": 0.1301122009754181,
      "learning_rate": 9.157348061512726e-07,
      "loss": 0.0326,
      "num_tokens": 39281955.0,
      "reward": 0.19618262629956007,
      "reward_std": 0.3902331814169884,
      "rewards/cosine_scaled_reward": 0.1961826179176569,
      "rewards/format_reward": 0.0,
      "step": 53
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 820.4486923217773,
      "epoch": 0.5475285171102662,
      "grad_norm": 0.12074018269777298,
      "learning_rate": 9.107103875602458e-07,
      "loss": -0.0192,
      "num_tokens": 40135629.0,
      "reward": 0.1222395608201623,
      "reward_std": 0.4181748628616333,
      "rewards/cosine_scaled_reward": 0.12223955010995269,
      "rewards/format_reward": 0.0,
      "step": 54
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 758.4922256469727,
      "epoch": 0.5576679340937896,
      "grad_norm": 0.14272169768810272,
      "learning_rate": 9.055551114853295e-07,
      "loss": 0.0269,
      "num_tokens": 40945630.0,
      "reward": 0.1248078744392842,
      "reward_std": 0.445852842181921,
      "rewards/cosine_scaled_reward": 0.12480786850210279,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 661.1495895385742,
      "epoch": 0.5678073510773131,
      "grad_norm": 0.2661519944667816,
      "learning_rate": 9.002706204621802e-07,
      "loss": 0.0013,
      "num_tokens": 41661700.0,
      "reward": 0.21717188507318497,
      "reward_std": 0.46877430006861687,
      "rewards/cosine_scaled_reward": 0.21717186993919313,
      "rewards/format_reward": 0.0,
      "step": 56
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 708.866096496582,
      "epoch": 0.5779467680608364,
      "grad_norm": 0.15228039026260376,
      "learning_rate": 8.948585981959578e-07,
      "loss": 0.0119,
      "num_tokens": 42428428.0,
      "reward": 0.25196343823336065,
      "reward_std": 0.3885682690888643,
      "rewards/cosine_scaled_reward": 0.2519634268246591,
      "rewards/format_reward": 0.0,
      "step": 57
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 707.9698944091797,
      "epoch": 0.5880861850443599,
      "grad_norm": 0.13880610466003418,
      "learning_rate": 8.893207690248775e-07,
      "loss": 0.0344,
      "num_tokens": 43183833.0,
      "reward": 0.22749723307788372,
      "reward_std": 0.41020799428224564,
      "rewards/cosine_scaled_reward": 0.22749722423031926,
      "rewards/format_reward": 0.0,
      "step": 58
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 698.7991333007812,
      "epoch": 0.5982256020278834,
      "grad_norm": 0.14641599357128143,
      "learning_rate": 8.836588973708128e-07,
      "loss": -0.0323,
      "num_tokens": 43951549.0,
      "reward": 0.11083738692104816,
      "reward_std": 0.45132191106677055,
      "rewards/cosine_scaled_reward": 0.11083737714216113,
      "rewards/format_reward": 0.0,
      "step": 59
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 749.3850860595703,
      "epoch": 0.6083650190114068,
      "grad_norm": 0.14265136420726776,
      "learning_rate": 8.778747871771291e-07,
      "loss": 0.0228,
      "num_tokens": 44760246.0,
      "reward": 0.15983119537122548,
      "reward_std": 0.4386202208697796,
      "rewards/cosine_scaled_reward": 0.15983118978329003,
      "rewards/format_reward": 0.0,
      "step": 60
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 700.7500305175781,
      "epoch": 0.6185044359949303,
      "grad_norm": 0.131936714053154,
      "learning_rate": 8.719702813339247e-07,
      "loss": 0.0532,
      "num_tokens": 45510462.0,
      "reward": 0.23273684654850513,
      "reward_std": 0.4278288222849369,
      "rewards/cosine_scaled_reward": 0.23273683816660196,
      "rewards/format_reward": 0.0,
      "step": 61
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 729.8102951049805,
      "epoch": 0.6286438529784537,
      "grad_norm": 0.13723866641521454,
      "learning_rate": 8.659472610908627e-07,
      "loss": -0.0047,
      "num_tokens": 46293764.0,
      "reward": 0.15899321692995727,
      "reward_std": 0.3875345978885889,
      "rewards/cosine_scaled_reward": 0.15899321576580405,
      "rewards/format_reward": 0.0,
      "step": 62
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 631.8594055175781,
      "epoch": 0.6387832699619772,
      "grad_norm": 0.15783964097499847,
      "learning_rate": 8.598076454577814e-07,
      "loss": 0.0349,
      "num_tokens": 46984206.0,
      "reward": 0.28133094896293187,
      "reward_std": 0.4228518716990948,
      "rewards/cosine_scaled_reward": 0.2813309368539194,
      "rewards/format_reward": 0.0,
      "step": 63
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 649.584846496582,
      "epoch": 0.6489226869455006,
      "grad_norm": 0.15222658216953278,
      "learning_rate": 8.535533905932737e-07,
      "loss": 0.0307,
      "num_tokens": 47714362.0,
      "reward": 0.28929764311760664,
      "reward_std": 0.3426767401397228,
      "rewards/cosine_scaled_reward": 0.2892976254224777,
      "rewards/format_reward": 0.0,
      "step": 64
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 679.0524826049805,
      "epoch": 0.6590621039290241,
      "grad_norm": 0.17544390261173248,
      "learning_rate": 8.471864891814304e-07,
      "loss": 0.0596,
      "num_tokens": 48441897.0,
      "reward": 0.2452558323275298,
      "reward_std": 0.4179961830377579,
      "rewards/cosine_scaled_reward": 0.24525582510977983,
      "rewards/format_reward": 0.0,
      "step": 65
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 734.794677734375,
      "epoch": 0.6692015209125475,
      "grad_norm": 0.13531185686588287,
      "learning_rate": 8.407089697969456e-07,
      "loss": 0.0013,
      "num_tokens": 49225137.0,
      "reward": 0.24510841415030882,
      "reward_std": 0.40535635873675346,
      "rewards/cosine_scaled_reward": 0.24510839313734323,
      "rewards/format_reward": 0.0,
      "step": 66
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 763.2321701049805,
      "epoch": 0.679340937896071,
      "grad_norm": 0.1570795476436615,
      "learning_rate": 8.341228962587881e-07,
      "loss": 0.0186,
      "num_tokens": 50045121.0,
      "reward": 0.20169485360383987,
      "reward_std": 0.37343652732670307,
      "rewards/cosine_scaled_reward": 0.2016948414966464,
      "rewards/format_reward": 0.0,
      "step": 67
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 628.7924346923828,
      "epoch": 0.6894803548795945,
      "grad_norm": 0.15324285626411438,
      "learning_rate": 8.274303669726426e-07,
      "loss": 0.0192,
      "num_tokens": 50729839.0,
      "reward": 0.32515534153208137,
      "reward_std": 0.39401572197675705,
      "rewards/cosine_scaled_reward": 0.32515532383695245,
      "rewards/format_reward": 0.0,
      "step": 68
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 743.4654388427734,
      "epoch": 0.6996197718631179,
      "grad_norm": 0.14744603633880615,
      "learning_rate": 8.206335142623304e-07,
      "loss": -0.0119,
      "num_tokens": 51525744.0,
      "reward": 0.1557501032948494,
      "reward_std": 0.4095492400228977,
      "rewards/cosine_scaled_reward": 0.15575009072199464,
      "rewards/format_reward": 0.0,
      "step": 69
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 700.7924575805664,
      "epoch": 0.7097591888466414,
      "grad_norm": 0.13947679102420807,
      "learning_rate": 8.137345036904259e-07,
      "loss": 0.0206,
      "num_tokens": 52273030.0,
      "reward": 0.14582239514857065,
      "reward_std": 0.4498204067349434,
      "rewards/cosine_scaled_reward": 0.14582238957518712,
      "rewards/format_reward": 0.0,
      "step": 70
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 674.7065048217773,
      "epoch": 0.7198986058301647,
      "grad_norm": 0.14676493406295776,
      "learning_rate": 8.067355333682797e-07,
      "loss": 0.0221,
      "num_tokens": 53016335.0,
      "reward": 0.18560006003826857,
      "reward_std": 0.39886176586151123,
      "rewards/cosine_scaled_reward": 0.1856000553816557,
      "rewards/format_reward": 0.0,
      "step": 71
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 669.2221298217773,
      "epoch": 0.7300380228136882,
      "grad_norm": 0.13405963778495789,
      "learning_rate": 7.996388332556734e-07,
      "loss": 0.0196,
      "num_tokens": 53744678.0,
      "reward": 0.22391042113304138,
      "reward_std": 0.40190940350294113,
      "rewards/cosine_scaled_reward": 0.2239104132167995,
      "rewards/format_reward": 0.0,
      "step": 72
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 718.1372909545898,
      "epoch": 0.7401774397972116,
      "grad_norm": 0.3063336908817291,
      "learning_rate": 7.924466644503264e-07,
      "loss": -0.0106,
      "num_tokens": 54520625.0,
      "reward": 0.20024659007322043,
      "reward_std": 0.42064137011766434,
      "rewards/cosine_scaled_reward": 0.2002465761033818,
      "rewards/format_reward": 0.0,
      "step": 73
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 641.3616333007812,
      "epoch": 0.7503168567807351,
      "grad_norm": 0.16075171530246735,
      "learning_rate": 7.85161318467482e-07,
      "loss": 0.044,
      "num_tokens": 55230669.0,
      "reward": 0.2528219725936651,
      "reward_std": 0.4147798642516136,
      "rewards/cosine_scaled_reward": 0.2528219521045685,
      "rewards/format_reward": 0.0,
      "step": 74
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 631.3672180175781,
      "epoch": 0.7604562737642585,
      "grad_norm": 0.6628883481025696,
      "learning_rate": 7.777851165098011e-07,
      "loss": 0.0325,
      "num_tokens": 55927054.0,
      "reward": 0.3146879682317376,
      "reward_std": 0.39327000454068184,
      "rewards/cosine_scaled_reward": 0.3146879549603909,
      "rewards/format_reward": 0.0,
      "step": 75
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 611.7455635070801,
      "epoch": 0.770595690747782,
      "grad_norm": 0.15209974348545074,
      "learning_rate": 7.703204087277988e-07,
      "loss": 0.0055,
      "num_tokens": 56602418.0,
      "reward": 0.135146701708436,
      "reward_std": 0.4354245401918888,
      "rewards/cosine_scaled_reward": 0.13514669938012958,
      "rewards/format_reward": 0.0,
      "step": 76
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 662.4486885070801,
      "epoch": 0.7807351077313055,
      "grad_norm": 0.14389804005622864,
      "learning_rate": 7.627695734710564e-07,
      "loss": -0.004,
      "num_tokens": 57318196.0,
      "reward": 0.24701710138469934,
      "reward_std": 0.3370856884866953,
      "rewards/cosine_scaled_reward": 0.2470170920714736,
      "rewards/format_reward": 0.0,
      "step": 77
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 684.2477912902832,
      "epoch": 0.7908745247148289,
      "grad_norm": 0.14715959131717682,
      "learning_rate": 7.551350165304499e-07,
      "loss": 0.0548,
      "num_tokens": 58049058.0,
      "reward": 0.30690577276982367,
      "reward_std": 0.41454331763088703,
      "rewards/cosine_scaled_reward": 0.30690575792687014,
      "rewards/format_reward": 0.0,
      "step": 78
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 651.327033996582,
      "epoch": 0.8010139416983524,
      "grad_norm": 0.13967670500278473,
      "learning_rate": 7.474191703716338e-07,
      "loss": 0.0039,
      "num_tokens": 58763207.0,
      "reward": 0.24645255785435438,
      "reward_std": 0.36919011175632477,
      "rewards/cosine_scaled_reward": 0.2464525457471609,
      "rewards/format_reward": 0.0,
      "step": 79
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 620.6060562133789,
      "epoch": 0.8111533586818758,
      "grad_norm": 0.1529722511768341,
      "learning_rate": 7.396244933600284e-07,
      "loss": 0.0101,
      "num_tokens": 59443558.0,
      "reward": 0.25752069521695375,
      "reward_std": 0.4100553132593632,
      "rewards/cosine_scaled_reward": 0.25752068776637316,
      "rewards/format_reward": 0.0,
      "step": 80
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 641.4230117797852,
      "epoch": 0.8212927756653993,
      "grad_norm": 0.1544528305530548,
      "learning_rate": 7.317534689775527e-07,
      "loss": 0.027,
      "num_tokens": 60137977.0,
      "reward": 0.22499833535403013,
      "reward_std": 0.406455984339118,
      "rewards/cosine_scaled_reward": 0.22499832697212696,
      "rewards/format_reward": 0.0,
      "step": 81
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 616.1852912902832,
      "epoch": 0.8314321926489227,
      "grad_norm": 0.20284396409988403,
      "learning_rate": 7.238086050313562e-07,
      "loss": -0.0023,
      "num_tokens": 60825335.0,
      "reward": 0.1846146322786808,
      "reward_std": 0.42002810910344124,
      "rewards/cosine_scaled_reward": 0.1846146178431809,
      "rewards/format_reward": 0.0,
      "step": 82
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 635.2466812133789,
      "epoch": 0.8415716096324461,
      "grad_norm": 0.163020059466362,
      "learning_rate": 7.157924328548002e-07,
      "loss": 0.0196,
      "num_tokens": 61528108.0,
      "reward": 0.2309768390841782,
      "reward_std": 0.4035301711410284,
      "rewards/cosine_scaled_reward": 0.2309768325649202,
      "rewards/format_reward": 0.0,
      "step": 83
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 632.155158996582,
      "epoch": 0.8517110266159695,
      "grad_norm": 0.15208573639392853,
      "learning_rate": 7.077075065009433e-07,
      "loss": 0.0356,
      "num_tokens": 62227831.0,
      "reward": 0.29583334969356656,
      "reward_std": 0.4564223028719425,
      "rewards/cosine_scaled_reward": 0.29583333851769567,
      "rewards/format_reward": 0.0,
      "step": 84
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 626.7265930175781,
      "epoch": 0.861850443599493,
      "grad_norm": 0.15748122334480286,
      "learning_rate": 6.995564019287869e-07,
      "loss": 0.0,
      "num_tokens": 62919194.0,
      "reward": 0.16931751649826765,
      "reward_std": 0.3771196622401476,
      "rewards/cosine_scaled_reward": 0.1693175071850419,
      "rewards/format_reward": 0.0,
      "step": 85
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 665.608283996582,
      "epoch": 0.8719898605830165,
      "grad_norm": 0.18250492215156555,
      "learning_rate": 6.913417161825449e-07,
      "loss": -0.0308,
      "num_tokens": 63650491.0,
      "reward": 0.20493324054405093,
      "reward_std": 0.3730998281389475,
      "rewards/cosine_scaled_reward": 0.2049332265742123,
      "rewards/format_reward": 0.0,
      "step": 86
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 622.6272583007812,
      "epoch": 0.8821292775665399,
      "grad_norm": 0.15512950718402863,
      "learning_rate": 6.830660665641897e-07,
      "loss": 0.0071,
      "num_tokens": 64338085.0,
      "reward": 0.2438387837028131,
      "reward_std": 0.40685577876865864,
      "rewards/cosine_scaled_reward": 0.24383877718355507,
      "rewards/format_reward": 0.0,
      "step": 87
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 664.6975784301758,
      "epoch": 0.8922686945500634,
      "grad_norm": 0.9565799832344055,
      "learning_rate": 6.747320897995492e-07,
      "loss": 0.0461,
      "num_tokens": 65062646.0,
      "reward": 0.25703565403819084,
      "reward_std": 0.4041633568704128,
      "rewards/cosine_scaled_reward": 0.2570356400683522,
      "rewards/format_reward": 0.0,
      "step": 88
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 588.4866333007812,
      "epoch": 0.9024081115335868,
      "grad_norm": 0.15928997099399567,
      "learning_rate": 6.66342441198212e-07,
      "loss": 0.003,
      "num_tokens": 65713834.0,
      "reward": 0.22450890811160207,
      "reward_std": 0.3792395293712616,
      "rewards/cosine_scaled_reward": 0.22450890205800533,
      "rewards/format_reward": 0.0,
      "step": 89
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 617.6417694091797,
      "epoch": 0.9125475285171103,
      "grad_norm": 0.16051417589187622,
      "learning_rate": 6.578997938075125e-07,
      "loss": 0.0237,
      "num_tokens": 66391713.0,
      "reward": 0.17285537254065275,
      "reward_std": 0.4654933251440525,
      "rewards/cosine_scaled_reward": 0.17285536974668503,
      "rewards/format_reward": 0.0,
      "step": 90
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 612.7511444091797,
      "epoch": 0.9226869455006337,
      "grad_norm": 0.1583762913942337,
      "learning_rate": 6.494068375608646e-07,
      "loss": -0.022,
      "num_tokens": 67075362.0,
      "reward": 0.2817280013114214,
      "reward_std": 0.3609468787908554,
      "rewards/cosine_scaled_reward": 0.28172798082232475,
      "rewards/format_reward": 0.0,
      "step": 91
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 647.9185485839844,
      "epoch": 0.9328263624841572,
      "grad_norm": 0.16609789431095123,
      "learning_rate": 6.408662784207149e-07,
      "loss": 0.0336,
      "num_tokens": 67795273.0,
      "reward": 0.2746627312153578,
      "reward_std": 0.41118355095386505,
      "rewards/cosine_scaled_reward": 0.2746627163141966,
      "rewards/format_reward": 0.0,
      "step": 92
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 586.2076110839844,
      "epoch": 0.9429657794676806,
      "grad_norm": 0.17962822318077087,
      "learning_rate": 6.322808375163895e-07,
      "loss": 0.0304,
      "num_tokens": 68447395.0,
      "reward": 0.2662193002179265,
      "reward_std": 0.4135792311280966,
      "rewards/cosine_scaled_reward": 0.2662192853167653,
      "rewards/format_reward": 0.0,
      "step": 93
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 628.2835159301758,
      "epoch": 0.9531051964512041,
      "grad_norm": 0.15532483160495758,
      "learning_rate": 6.236532502771077e-07,
      "loss": -0.0041,
      "num_tokens": 69143937.0,
      "reward": 0.22241216152906418,
      "reward_std": 0.41293996572494507,
      "rewards/cosine_scaled_reward": 0.22241215035319328,
      "rewards/format_reward": 0.0,
      "step": 94
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 581.4933280944824,
      "epoch": 0.9632446134347274,
      "grad_norm": 0.15709497034549713,
      "learning_rate": 6.149862655604403e-07,
      "loss": 0.0438,
      "num_tokens": 69790643.0,
      "reward": 0.2721885642968118,
      "reward_std": 0.40717606246471405,
      "rewards/cosine_scaled_reward": 0.2721885568462312,
      "rewards/format_reward": 0.0,
      "step": 95
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 614.9408721923828,
      "epoch": 0.973384030418251,
      "grad_norm": 0.16992299258708954,
      "learning_rate": 6.062826447764883e-07,
      "loss": 0.037,
      "num_tokens": 70464374.0,
      "reward": 0.2846832238137722,
      "reward_std": 0.3609850201755762,
      "rewards/cosine_scaled_reward": 0.2846832163631916,
      "rewards/format_reward": 0.0,
      "step": 96
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 633.0982398986816,
      "epoch": 0.9835234474017744,
      "grad_norm": 0.14964738488197327,
      "learning_rate": 5.975451610080642e-07,
      "loss": 0.0147,
      "num_tokens": 71159438.0,
      "reward": 0.19606254808604717,
      "reward_std": 0.3826338779181242,
      "rewards/cosine_scaled_reward": 0.19606254203245044,
      "rewards/format_reward": 0.0,
      "step": 97
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 603.615234375,
      "epoch": 0.9936628643852978,
      "grad_norm": 0.16081441938877106,
      "learning_rate": 5.887765981271517e-07,
      "loss": 0.0428,
      "num_tokens": 71852378.0,
      "reward": 0.2125524217262864,
      "reward_std": 0.3838059641420841,
      "rewards/cosine_scaled_reward": 0.21255241055041552,
      "rewards/format_reward": 0.0,
      "step": 98
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 590.7076225280762,
      "epoch": 1.0101394169835234,
      "grad_norm": 0.1603391319513321,
      "learning_rate": 5.7997974990793e-07,
      "loss": 0.0057,
      "num_tokens": 72510644.0,
      "reward": 0.2637674935394898,
      "reward_std": 0.4375726915895939,
      "rewards/cosine_scaled_reward": 0.26376748620532453,
      "rewards/format_reward": 0.0,
      "step": 99
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 611.1406593322754,
      "epoch": 1.020278833967047,
      "grad_norm": 0.16759544610977173,
      "learning_rate": 5.711574191366427e-07,
      "loss": 0.0233,
      "num_tokens": 73185602.0,
      "reward": 0.24167153984308243,
      "reward_std": 0.3558240346610546,
      "rewards/cosine_scaled_reward": 0.2416715300641954,
      "rewards/format_reward": 0.0,
      "step": 100
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 630.9598426818848,
      "epoch": 1.0304182509505704,
      "grad_norm": 0.16307219862937927,
      "learning_rate": 5.623124167185929e-07,
      "loss": 0.0067,
      "num_tokens": 73873470.0,
      "reward": 0.26334861340001225,
      "reward_std": 0.3806659858673811,
      "rewards/cosine_scaled_reward": 0.26334861130453646,
      "rewards/format_reward": 0.0,
      "step": 101
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 563.8035888671875,
      "epoch": 1.0405576679340938,
      "grad_norm": 0.18253548443317413,
      "learning_rate": 5.534475607825565e-07,
      "loss": 0.0047,
      "num_tokens": 74505926.0,
      "reward": 0.24372864328324795,
      "reward_std": 0.3952293395996094,
      "rewards/cosine_scaled_reward": 0.2437286302447319,
      "rewards/format_reward": 0.0,
      "step": 102
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 634.3750267028809,
      "epoch": 1.0506970849176172,
      "grad_norm": 0.16632695496082306,
      "learning_rate": 5.445656757828879e-07,
      "loss": -0.0255,
      "num_tokens": 75205630.0,
      "reward": 0.07845631497912109,
      "reward_std": 0.38748307153582573,
      "rewards/cosine_scaled_reward": 0.07845630997326225,
      "rewards/format_reward": 0.0,
      "step": 103
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.0971221923828,
      "epoch": 1.0608365019011408,
      "grad_norm": 0.14292435348033905,
      "learning_rate": 5.356695915996161e-07,
      "loss": -0.0234,
      "num_tokens": 75878573.0,
      "reward": 0.1749452566727996,
      "reward_std": 0.44058991968631744,
      "rewards/cosine_scaled_reward": 0.17494524829089642,
      "rewards/format_reward": 0.0,
      "step": 104
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 558.7801628112793,
      "epoch": 1.0709759188846641,
      "grad_norm": 0.16877137124538422,
      "learning_rate": 5.267621426368075e-07,
      "loss": 0.0282,
      "num_tokens": 76498456.0,
      "reward": 0.29683027416467667,
      "reward_std": 0.43279460445046425,
      "rewards/cosine_scaled_reward": 0.2968302555382252,
      "rewards/format_reward": 0.0,
      "step": 105
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 506.5335006713867,
      "epoch": 1.0811153358681875,
      "grad_norm": 0.17591001093387604,
      "learning_rate": 5.178461669194903e-07,
      "loss": 0.0483,
      "num_tokens": 77079518.0,
      "reward": 0.3861896749585867,
      "reward_std": 0.3774683568626642,
      "rewards/cosine_scaled_reward": 0.3861896488815546,
      "rewards/format_reward": 0.0,
      "step": 106
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 534.6607398986816,
      "epoch": 1.091254752851711,
      "grad_norm": 0.15835636854171753,
      "learning_rate": 5.08924505189423e-07,
      "loss": -0.0117,
      "num_tokens": 77684534.0,
      "reward": 0.28510209172964096,
      "reward_std": 0.31983859837055206,
      "rewards/cosine_scaled_reward": 0.28510207310318947,
      "rewards/format_reward": 0.0,
      "step": 107
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 594.6439895629883,
      "epoch": 1.1013941698352345,
      "grad_norm": 0.20408962666988373,
      "learning_rate": 5e-07,
      "loss": 0.0081,
      "num_tokens": 78341455.0,
      "reward": 0.2976014781743288,
      "reward_std": 0.3984498456120491,
      "rewards/cosine_scaled_reward": 0.29760145768523216,
      "rewards/format_reward": 0.0,
      "step": 108
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 630.444221496582,
      "epoch": 1.111533586818758,
      "grad_norm": 0.1533069759607315,
      "learning_rate": 4.91075494810577e-07,
      "loss": 0.0265,
      "num_tokens": 79024325.0,
      "reward": 0.25680719036608934,
      "reward_std": 0.4040773529559374,
      "rewards/cosine_scaled_reward": 0.25680717546492815,
      "rewards/format_reward": 0.0,
      "step": 109
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 637.928596496582,
      "epoch": 1.1216730038022813,
      "grad_norm": 0.16302607953548431,
      "learning_rate": 4.821538330805098e-07,
      "loss": -0.0033,
      "num_tokens": 79722429.0,
      "reward": 0.2886841148138046,
      "reward_std": 0.3645545728504658,
      "rewards/cosine_scaled_reward": 0.28868410736322403,
      "rewards/format_reward": 0.0,
      "step": 110
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 614.0390968322754,
      "epoch": 1.131812420785805,
      "grad_norm": 0.15493924915790558,
      "learning_rate": 4.732378573631924e-07,
      "loss": 0.0113,
      "num_tokens": 80395336.0,
      "reward": 0.25373244076035917,
      "reward_std": 0.3782533872872591,
      "rewards/cosine_scaled_reward": 0.2537324334261939,
      "rewards/format_reward": 0.0,
      "step": 111
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 615.8136520385742,
      "epoch": 1.1419518377693283,
      "grad_norm": 1.329131841659546,
      "learning_rate": 4.643304084003838e-07,
      "loss": 0.028,
      "num_tokens": 81074553.0,
      "reward": 0.19473237730562687,
      "reward_std": 0.4509877860546112,
      "rewards/cosine_scaled_reward": 0.19473236822523177,
      "rewards/format_reward": 0.0,
      "step": 112
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 605.2098503112793,
      "epoch": 1.1520912547528517,
      "grad_norm": 0.16996121406555176,
      "learning_rate": 4.55434324217112e-07,
      "loss": 0.0026,
      "num_tokens": 81746061.0,
      "reward": 0.1302806935273111,
      "reward_std": 0.38874428533017635,
      "rewards/cosine_scaled_reward": 0.13028068700805306,
      "rewards/format_reward": 0.0,
      "step": 113
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 596.4888648986816,
      "epoch": 1.162230671736375,
      "grad_norm": 0.1666853278875351,
      "learning_rate": 4.4655243921744367e-07,
      "loss": -0.0409,
      "num_tokens": 82405475.0,
      "reward": 0.21928726602345705,
      "reward_std": 0.3403788786381483,
      "rewards/cosine_scaled_reward": 0.21928725019097328,
      "rewards/format_reward": 0.0,
      "step": 114
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 564.8895378112793,
      "epoch": 1.1723700887198987,
      "grad_norm": 0.16143201291561127,
      "learning_rate": 4.37687583281407e-07,
      "loss": -0.0382,
      "num_tokens": 83038832.0,
      "reward": 0.21677181296399795,
      "reward_std": 0.35355726443231106,
      "rewards/cosine_scaled_reward": 0.21677180923870765,
      "rewards/format_reward": 0.0,
      "step": 115
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 553.4587287902832,
      "epoch": 1.182509505703422,
      "grad_norm": 0.16877496242523193,
      "learning_rate": 4.2884258086335745e-07,
      "loss": 0.0313,
      "num_tokens": 83673971.0,
      "reward": 0.2728145425207913,
      "reward_std": 0.42727693170309067,
      "rewards/cosine_scaled_reward": 0.2728145350702107,
      "rewards/format_reward": 0.0,
      "step": 116
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 550.5703315734863,
      "epoch": 1.1926489226869454,
      "grad_norm": 0.17830805480480194,
      "learning_rate": 4.2002025009206987e-07,
      "loss": 0.0008,
      "num_tokens": 84299234.0,
      "reward": 0.1771111206180649,
      "reward_std": 0.42238056659698486,
      "rewards/cosine_scaled_reward": 0.17711111830431037,
      "rewards/format_reward": 0.0,
      "step": 117
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 588.3515815734863,
      "epoch": 1.202788339670469,
      "grad_norm": 0.17095865309238434,
      "learning_rate": 4.1122340187284845e-07,
      "loss": 0.0134,
      "num_tokens": 84948685.0,
      "reward": 0.20526447892189026,
      "reward_std": 0.43867457658052444,
      "rewards/cosine_scaled_reward": 0.20526447147130966,
      "rewards/format_reward": 0.0,
      "step": 118
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 536.7544975280762,
      "epoch": 1.2129277566539924,
      "grad_norm": 0.1749478578567505,
      "learning_rate": 4.0245483899193586e-07,
      "loss": 0.0174,
      "num_tokens": 85556993.0,
      "reward": 0.24879271956160665,
      "reward_std": 0.38236525282263756,
      "rewards/cosine_scaled_reward": 0.24879270186647773,
      "rewards/format_reward": 0.0,
      "step": 119
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 558.5870742797852,
      "epoch": 1.2230671736375158,
      "grad_norm": 1.2731257677078247,
      "learning_rate": 3.937173552235116e-07,
      "loss": -0.0515,
      "num_tokens": 86190295.0,
      "reward": 0.1734993983991444,
      "reward_std": 0.3362387754023075,
      "rewards/cosine_scaled_reward": 0.1734993918798864,
      "rewards/format_reward": 0.0,
      "step": 120
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 564.9698867797852,
      "epoch": 1.2332065906210392,
      "grad_norm": 0.20518043637275696,
      "learning_rate": 3.850137344395598e-07,
      "loss": 0.008,
      "num_tokens": 86819724.0,
      "reward": 0.2829995099455118,
      "reward_std": 0.33430546894669533,
      "rewards/cosine_scaled_reward": 0.2829994931817055,
      "rewards/format_reward": 0.0,
      "step": 121
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 628.7623138427734,
      "epoch": 1.2433460076045628,
      "grad_norm": 0.1478431224822998,
      "learning_rate": 3.763467497228922e-07,
      "loss": -0.0359,
      "num_tokens": 87521127.0,
      "reward": 0.16334752831608057,
      "reward_std": 0.3575255237519741,
      "rewards/cosine_scaled_reward": 0.16334752179682255,
      "rewards/format_reward": 0.0,
      "step": 122
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 576.7277030944824,
      "epoch": 1.2534854245880862,
      "grad_norm": 0.20034752786159515,
      "learning_rate": 3.677191624836106e-07,
      "loss": 0.0205,
      "num_tokens": 88159587.0,
      "reward": 0.32477190531790257,
      "reward_std": 0.3922874592244625,
      "rewards/cosine_scaled_reward": 0.3247718848288059,
      "rewards/format_reward": 0.0,
      "step": 123
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 611.7667617797852,
      "epoch": 1.2636248415716096,
      "grad_norm": 0.1693291813135147,
      "learning_rate": 3.591337215792851e-07,
      "loss": 0.0109,
      "num_tokens": 88835898.0,
      "reward": 0.23217581922654063,
      "reward_std": 0.34129480831325054,
      "rewards/cosine_scaled_reward": 0.23217580071650445,
      "rewards/format_reward": 0.0,
      "step": 124
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 557.9799385070801,
      "epoch": 1.2737642585551332,
      "grad_norm": 0.1676970273256302,
      "learning_rate": 3.505931624391355e-07,
      "loss": 0.0225,
      "num_tokens": 89454112.0,
      "reward": 0.20786370153655298,
      "reward_std": 0.4651025366038084,
      "rewards/cosine_scaled_reward": 0.20786369033157825,
      "rewards/format_reward": 0.0,
      "step": 125
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 545.6931037902832,
      "epoch": 1.2839036755386566,
      "grad_norm": 0.1678163707256317,
      "learning_rate": 3.421002061924876e-07,
      "loss": 0.0151,
      "num_tokens": 90069917.0,
      "reward": 0.2275780588388443,
      "reward_std": 0.4175031315535307,
      "rewards/cosine_scaled_reward": 0.22757804207503796,
      "rewards/format_reward": 0.0,
      "step": 126
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 540.5000152587891,
      "epoch": 1.29404309252218,
      "grad_norm": 0.17660397291183472,
      "learning_rate": 3.3365755880178807e-07,
      "loss": 0.0301,
      "num_tokens": 90675909.0,
      "reward": 0.2635859549045563,
      "reward_std": 0.41772597655653954,
      "rewards/cosine_scaled_reward": 0.2635859474539757,
      "rewards/format_reward": 0.0,
      "step": 127
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 594.244441986084,
      "epoch": 1.3041825095057034,
      "grad_norm": 0.17012734711170197,
      "learning_rate": 3.2526791020045087e-07,
      "loss": 0.0215,
      "num_tokens": 91340952.0,
      "reward": 0.149813212454319,
      "reward_std": 0.4470374099910259,
      "rewards/cosine_scaled_reward": 0.14981320640072227,
      "rewards/format_reward": 0.0,
      "step": 128
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 542.3348503112793,
      "epoch": 1.3143219264892267,
      "grad_norm": 0.174094557762146,
      "learning_rate": 3.169339334358104e-07,
      "loss": 0.0523,
      "num_tokens": 91960228.0,
      "reward": 0.26255749678239226,
      "reward_std": 0.32982037030160427,
      "rewards/cosine_scaled_reward": 0.26255749305710196,
      "rewards/format_reward": 0.0,
      "step": 129
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 567.3292694091797,
      "epoch": 1.3244613434727504,
      "grad_norm": 0.16861200332641602,
      "learning_rate": 3.086582838174551e-07,
      "loss": -0.0835,
      "num_tokens": 92601043.0,
      "reward": 0.1493756880190631,
      "reward_std": 0.3699100259691477,
      "rewards/cosine_scaled_reward": 0.1493756801019117,
      "rewards/format_reward": 0.0,
      "step": 130
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 578.1283683776855,
      "epoch": 1.3346007604562737,
      "grad_norm": 1.8984020948410034,
      "learning_rate": 3.004435980712129e-07,
      "loss": -0.0221,
      "num_tokens": 93242510.0,
      "reward": 0.23330406891182065,
      "reward_std": 0.4134122505784035,
      "rewards/cosine_scaled_reward": 0.23330405401065946,
      "rewards/format_reward": 0.0,
      "step": 131
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 538.444221496582,
      "epoch": 1.3447401774397973,
      "grad_norm": 0.1705993413925171,
      "learning_rate": 2.922924934990568e-07,
      "loss": 0.0105,
      "num_tokens": 93850108.0,
      "reward": 0.18292724061757326,
      "reward_std": 0.4480227939784527,
      "rewards/cosine_scaled_reward": 0.18292723130434752,
      "rewards/format_reward": 0.0,
      "step": 132
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 524.5457763671875,
      "epoch": 1.3548795944233207,
      "grad_norm": 0.20817777514457703,
      "learning_rate": 2.8420756714519954e-07,
      "loss": 0.0035,
      "num_tokens": 94464405.0,
      "reward": 0.2242459380067885,
      "reward_std": 0.3740137480199337,
      "rewards/cosine_scaled_reward": 0.2242459305562079,
      "rewards/format_reward": 0.0,
      "step": 133
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 542.7221260070801,
      "epoch": 1.3650190114068441,
      "grad_norm": 0.1813177764415741,
      "learning_rate": 2.7619139496864376e-07,
      "loss": -0.029,
      "num_tokens": 95086668.0,
      "reward": 0.13100900989957154,
      "reward_std": 0.3776312284171581,
      "rewards/cosine_scaled_reward": 0.13100900826975703,
      "rewards/format_reward": 0.0,
      "step": 134
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 500.7243461608887,
      "epoch": 1.3751584283903675,
      "grad_norm": 0.18803931772708893,
      "learning_rate": 2.6824653102244727e-07,
      "loss": 0.0016,
      "num_tokens": 95669389.0,
      "reward": 0.189885966014117,
      "reward_std": 0.45219165831804276,
      "rewards/cosine_scaled_reward": 0.18988595390692353,
      "rewards/format_reward": 0.0,
      "step": 135
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 512.5524864196777,
      "epoch": 1.385297845373891,
      "grad_norm": 0.21896012127399445,
      "learning_rate": 2.603755066399718e-07,
      "loss": -0.0237,
      "num_tokens": 96258716.0,
      "reward": 0.17706315219402313,
      "reward_std": 0.4132867194712162,
      "rewards/cosine_scaled_reward": 0.1770631456747651,
      "rewards/format_reward": 0.0,
      "step": 136
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 611.787971496582,
      "epoch": 1.3954372623574145,
      "grad_norm": 0.16569751501083374,
      "learning_rate": 2.5258082962836614e-07,
      "loss": -0.0733,
      "num_tokens": 96941438.0,
      "reward": 0.07232440868392587,
      "reward_std": 0.3806895762681961,
      "rewards/cosine_scaled_reward": 0.07232440612278879,
      "rewards/format_reward": 0.0,
      "step": 137
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 532.1986846923828,
      "epoch": 1.4055766793409379,
      "grad_norm": 0.18526431918144226,
      "learning_rate": 2.4486498346955023e-07,
      "loss": 0.0031,
      "num_tokens": 97541120.0,
      "reward": 0.18689691158942878,
      "reward_std": 0.4354799725115299,
      "rewards/cosine_scaled_reward": 0.18689690087921917,
      "rewards/format_reward": 0.0,
      "step": 138
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 538.1942138671875,
      "epoch": 1.4157160963244613,
      "grad_norm": 0.1791687160730362,
      "learning_rate": 2.372304265289436e-07,
      "loss": -0.0459,
      "num_tokens": 98147638.0,
      "reward": 0.13275248603895307,
      "reward_std": 0.44364722445607185,
      "rewards/cosine_scaled_reward": 0.1327524788212031,
      "rewards/format_reward": 0.0,
      "step": 139
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 511.8605079650879,
      "epoch": 1.4258555133079849,
      "grad_norm": 0.1867491602897644,
      "learning_rate": 2.2967959127220137e-07,
      "loss": 0.0068,
      "num_tokens": 98728225.0,
      "reward": 0.3189204416703433,
      "reward_std": 0.3536707144230604,
      "rewards/cosine_scaled_reward": 0.3189204225782305,
      "rewards/format_reward": 0.0,
      "step": 140
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 527.9910926818848,
      "epoch": 1.4359949302915083,
      "grad_norm": 0.18118509650230408,
      "learning_rate": 2.2221488349019902e-07,
      "loss": -0.0149,
      "num_tokens": 99349665.0,
      "reward": 0.136872966773808,
      "reward_std": 0.42309295386075974,
      "rewards/cosine_scaled_reward": 0.13687296118587255,
      "rewards/format_reward": 0.0,
      "step": 141
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 508.416316986084,
      "epoch": 1.4461343472750317,
      "grad_norm": 0.17504200339317322,
      "learning_rate": 2.1483868153251788e-07,
      "loss": -0.0121,
      "num_tokens": 99936750.0,
      "reward": 0.24867849051952362,
      "reward_std": 0.4239983968436718,
      "rewards/cosine_scaled_reward": 0.2486784835346043,
      "rewards/format_reward": 0.0,
      "step": 142
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 571.9408798217773,
      "epoch": 1.456273764258555,
      "grad_norm": 0.2036023736000061,
      "learning_rate": 2.0755333554967346e-07,
      "loss": -0.0038,
      "num_tokens": 100586409.0,
      "reward": 0.11940343677997589,
      "reward_std": 0.5181355699896812,
      "rewards/cosine_scaled_reward": 0.11940343119204044,
      "rewards/format_reward": 0.0,
      "step": 143
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 599.2779235839844,
      "epoch": 1.4664131812420786,
      "grad_norm": 0.17052923142910004,
      "learning_rate": 2.0036116674432652e-07,
      "loss": -0.0245,
      "num_tokens": 101259810.0,
      "reward": 0.12587101384997368,
      "reward_std": 0.3964061960577965,
      "rewards/cosine_scaled_reward": 0.12587100733071566,
      "rewards/format_reward": 0.0,
      "step": 144
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 473.37948989868164,
      "epoch": 1.476552598225602,
      "grad_norm": 0.1972968876361847,
      "learning_rate": 1.9326446663172035e-07,
      "loss": -0.0245,
      "num_tokens": 101815438.0,
      "reward": 0.2772935745306313,
      "reward_std": 0.44544481486082077,
      "rewards/cosine_scaled_reward": 0.2772935628890991,
      "rewards/format_reward": 0.0,
      "step": 145
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 479.25560760498047,
      "epoch": 1.4866920152091254,
      "grad_norm": 0.236617311835289,
      "learning_rate": 1.8626549630957395e-07,
      "loss": 0.0277,
      "num_tokens": 102372995.0,
      "reward": 0.2525997795164585,
      "reward_std": 0.4284665212035179,
      "rewards/cosine_scaled_reward": 0.25259976275265217,
      "rewards/format_reward": 0.0,
      "step": 146
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 550.899585723877,
      "epoch": 1.496831432192649,
      "grad_norm": 0.17915715277194977,
      "learning_rate": 1.7936648573766954e-07,
      "loss": -0.0269,
      "num_tokens": 103003313.0,
      "reward": 0.16996209789067507,
      "reward_std": 0.42602597922086716,
      "rewards/cosine_scaled_reward": 0.1699620874132961,
      "rewards/format_reward": 0.0,
      "step": 147
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 564.1919822692871,
      "epoch": 1.5069708491761724,
      "grad_norm": 0.18789808452129364,
      "learning_rate": 1.725696330273575e-07,
      "loss": -0.0676,
      "num_tokens": 103639173.0,
      "reward": 0.16659816866740584,
      "reward_std": 0.36924581229686737,
      "rewards/cosine_scaled_reward": 0.16659815539605916,
      "rewards/format_reward": 0.0,
      "step": 148
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 487.6774711608887,
      "epoch": 1.5171102661596958,
      "grad_norm": 0.20230001211166382,
      "learning_rate": 1.65877103741212e-07,
      "loss": 0.0106,
      "num_tokens": 104200860.0,
      "reward": 0.2995794331654906,
      "reward_std": 0.4441289193928242,
      "rewards/cosine_scaled_reward": 0.2995794131420553,
      "rewards/format_reward": 0.0,
      "step": 149
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 515.7678756713867,
      "epoch": 1.5272496831432192,
      "grad_norm": 0.20191656053066254,
      "learning_rate": 1.592910302030544e-07,
      "loss": 0.0207,
      "num_tokens": 104782292.0,
      "reward": 0.17506056488491595,
      "reward_std": 0.4823239389806986,
      "rewards/cosine_scaled_reward": 0.17506056022830307,
      "rewards/format_reward": 0.0,
      "step": 150
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 517.0513648986816,
      "epoch": 1.5373891001267426,
      "grad_norm": 0.18813706934452057,
      "learning_rate": 1.5281351081856976e-07,
      "loss": 0.0014,
      "num_tokens": 105376826.0,
      "reward": 0.19976339329605253,
      "reward_std": 0.43244756758213043,
      "rewards/cosine_scaled_reward": 0.19976338025753648,
      "rewards/format_reward": 0.0,
      "step": 151
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 517.2968978881836,
      "epoch": 1.5475285171102662,
      "grad_norm": 0.18564638495445251,
      "learning_rate": 1.4644660940672627e-07,
      "loss": -0.0032,
      "num_tokens": 105965916.0,
      "reward": 0.29871877538971603,
      "reward_std": 0.39689321629703045,
      "rewards/cosine_scaled_reward": 0.29871875676326454,
      "rewards/format_reward": 0.0,
      "step": 152
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 506.71319580078125,
      "epoch": 1.5576679340937896,
      "grad_norm": 0.17704474925994873,
      "learning_rate": 1.4019235454221856e-07,
      "loss": -0.0185,
      "num_tokens": 106554003.0,
      "reward": 0.21053828485310078,
      "reward_std": 0.3987581580877304,
      "rewards/cosine_scaled_reward": 0.21053827833384275,
      "rewards/format_reward": 0.0,
      "step": 153
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 475.9431037902832,
      "epoch": 1.5678073510773132,
      "grad_norm": 0.19738495349884033,
      "learning_rate": 1.3405273890913737e-07,
      "loss": 0.0262,
      "num_tokens": 107102464.0,
      "reward": 0.21143945213407278,
      "reward_std": 0.42565931007266045,
      "rewards/cosine_scaled_reward": 0.21143944095820189,
      "rewards/format_reward": 0.0,
      "step": 154
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 508.6618537902832,
      "epoch": 1.5779467680608366,
      "grad_norm": 0.328647643327713,
      "learning_rate": 1.280297186660752e-07,
      "loss": 0.0234,
      "num_tokens": 107676601.0,
      "reward": 0.25387828424572945,
      "reward_std": 0.4254852868616581,
      "rewards/cosine_scaled_reward": 0.25387826934456825,
      "rewards/format_reward": 0.0,
      "step": 155
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 587.3850631713867,
      "epoch": 1.58808618504436,
      "grad_norm": 0.17507341504096985,
      "learning_rate": 1.2212521282287093e-07,
      "loss": -0.0302,
      "num_tokens": 108322826.0,
      "reward": 0.1246091895736754,
      "reward_std": 0.43387965485453606,
      "rewards/cosine_scaled_reward": 0.12460918258875608,
      "rewards/format_reward": 0.0,
      "step": 156
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 570.8884086608887,
      "epoch": 1.5982256020278833,
      "grad_norm": 0.17769315838813782,
      "learning_rate": 1.1634110262918717e-07,
      "loss": -0.0209,
      "num_tokens": 108978350.0,
      "reward": 0.17509674560278654,
      "reward_std": 0.3465359937399626,
      "rewards/cosine_scaled_reward": 0.17509673908352852,
      "rewards/format_reward": 0.0,
      "step": 157
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 546.3772621154785,
      "epoch": 1.6083650190114067,
      "grad_norm": 0.18788230419158936,
      "learning_rate": 1.1067923097512255e-07,
      "loss": -0.0358,
      "num_tokens": 109598616.0,
      "reward": 0.12754715513437986,
      "reward_std": 0.4496493525803089,
      "rewards/cosine_scaled_reward": 0.12754714622860774,
      "rewards/format_reward": 0.0,
      "step": 158
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 543.7288246154785,
      "epoch": 1.6185044359949303,
      "grad_norm": 0.2574346661567688,
      "learning_rate": 1.0514140180404202e-07,
      "loss": -0.0512,
      "num_tokens": 110214853.0,
      "reward": 0.08189231436699629,
      "reward_std": 0.41547736898064613,
      "rewards/cosine_scaled_reward": 0.08189231110736728,
      "rewards/format_reward": 0.0,
      "step": 159
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 607.2890892028809,
      "epoch": 1.6286438529784537,
      "grad_norm": 0.22311408817768097,
      "learning_rate": 9.972937953781984e-08,
      "loss": -0.0035,
      "num_tokens": 110905952.0,
      "reward": 0.07692541873257142,
      "reward_std": 0.4941936247050762,
      "rewards/cosine_scaled_reward": 0.07692541686992627,
      "rewards/format_reward": 0.0,
      "step": 160
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 528.2031402587891,
      "epoch": 1.6387832699619773,
      "grad_norm": 1.5113683938980103,
      "learning_rate": 9.444488851467041e-08,
      "loss": -0.0166,
      "num_tokens": 111496894.0,
      "reward": 0.23110088612884283,
      "reward_std": 0.44028190523386,
      "rewards/cosine_scaled_reward": 0.23110087541863322,
      "rewards/format_reward": 0.0,
      "step": 161
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 523.2132034301758,
      "epoch": 1.6489226869455007,
      "grad_norm": 0.18456611037254333,
      "learning_rate": 8.928961243975436e-08,
      "loss": -0.0148,
      "num_tokens": 112098549.0,
      "reward": 0.16438766405917704,
      "reward_std": 0.4540070556104183,
      "rewards/cosine_scaled_reward": 0.16713074478320777,
      "rewards/format_reward": 0.0,
      "step": 162
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 490.20092010498047,
      "epoch": 1.659062103929024,
      "grad_norm": 0.20410823822021484,
      "learning_rate": 8.426519384872732e-08,
      "loss": -0.0209,
      "num_tokens": 112659905.0,
      "reward": 0.2585650213295594,
      "reward_std": 0.42336183600127697,
      "rewards/cosine_scaled_reward": 0.25856500840745866,
      "rewards/format_reward": 0.0,
      "step": 163
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 515.3370704650879,
      "epoch": 1.6692015209125475,
      "grad_norm": 0.2762053906917572,
      "learning_rate": 7.937323358440934e-08,
      "loss": 0.036,
      "num_tokens": 113252679.0,
      "reward": 0.21294681075960398,
      "reward_std": 0.4499225113540888,
      "rewards/cosine_scaled_reward": 0.2129468061029911,
      "rewards/format_reward": 0.0,
      "step": 164
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 517.8939971923828,
      "epoch": 1.6793409378960709,
      "grad_norm": 0.18798772990703583,
      "learning_rate": 7.461529028673463e-08,
      "loss": 0.0349,
      "num_tokens": 113849520.0,
      "reward": 0.2390075612347573,
      "reward_std": 0.4681055396795273,
      "rewards/cosine_scaled_reward": 0.23900754831265658,
      "rewards/format_reward": 0.0,
      "step": 165
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 557.4988975524902,
      "epoch": 1.6894803548795945,
      "grad_norm": 0.19520263373851776,
      "learning_rate": 6.999287989614971e-08,
      "loss": 0.0068,
      "num_tokens": 114481479.0,
      "reward": 0.11839688033796847,
      "reward_std": 0.5270655192434788,
      "rewards/cosine_scaled_reward": 0.11839687242172658,
      "rewards/format_reward": 0.0,
      "step": 166
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 484.58261489868164,
      "epoch": 1.6996197718631179,
      "grad_norm": 0.21452556550502777,
      "learning_rate": 6.550747517061656e-08,
      "loss": 0.0029,
      "num_tokens": 115042209.0,
      "reward": 0.24680990353226662,
      "reward_std": 0.4576365016400814,
      "rewards/cosine_scaled_reward": 0.2468099007382989,
      "rewards/format_reward": 0.0,
      "step": 167
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 516.2667655944824,
      "epoch": 1.7097591888466415,
      "grad_norm": 0.18506427109241486,
      "learning_rate": 6.116050521637218e-08,
      "loss": 0.019,
      "num_tokens": 115633112.0,
      "reward": 0.22252054838463664,
      "reward_std": 0.4747694581747055,
      "rewards/cosine_scaled_reward": 0.22252054512500763,
      "rewards/format_reward": 0.0,
      "step": 168
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 511.237735748291,
      "epoch": 1.7198986058301649,
      "grad_norm": 0.19289632141590118,
      "learning_rate": 5.6953355032598795e-08,
      "loss": -0.0568,
      "num_tokens": 116228253.0,
      "reward": 0.08733667735941708,
      "reward_std": 0.49206770583987236,
      "rewards/cosine_scaled_reward": 0.08733667538035661,
      "rewards/format_reward": 0.0,
      "step": 169
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 520.1964530944824,
      "epoch": 1.7300380228136882,
      "grad_norm": 0.19966185092926025,
      "learning_rate": 5.288736507014435e-08,
      "loss": -0.0308,
      "num_tokens": 116818349.0,
      "reward": 0.20070420764386654,
      "reward_std": 0.45568813756108284,
      "rewards/cosine_scaled_reward": 0.20070419739931822,
      "rewards/format_reward": 0.0,
      "step": 170
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 482.51565170288086,
      "epoch": 1.7401774397972116,
      "grad_norm": 0.2040020376443863,
      "learning_rate": 4.896383080443933e-08,
      "loss": -0.0247,
      "num_tokens": 117375435.0,
      "reward": 0.13767189331701957,
      "reward_std": 0.42815835028886795,
      "rewards/cosine_scaled_reward": 0.13767188772908412,
      "rewards/format_reward": 0.0,
      "step": 171
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 504.1607437133789,
      "epoch": 1.750316856780735,
      "grad_norm": 0.20132046937942505,
      "learning_rate": 4.518400232274078e-08,
      "loss": -0.0345,
      "num_tokens": 117957667.0,
      "reward": 0.17123566241934896,
      "reward_std": 0.3756715804338455,
      "rewards/cosine_scaled_reward": 0.17370464373379946,
      "rewards/format_reward": 0.0,
      "step": 172
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 460.175235748291,
      "epoch": 1.7604562737642584,
      "grad_norm": 0.2837740480899811,
      "learning_rate": 4.1549083925840165e-08,
      "loss": -0.0105,
      "num_tokens": 118487432.0,
      "reward": 0.2758436817675829,
      "reward_std": 0.4501863084733486,
      "rewards/cosine_scaled_reward": 0.2758436668664217,
      "rewards/format_reward": 0.0,
      "step": 173
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 496.6093940734863,
      "epoch": 1.770595690747782,
      "grad_norm": 0.18853703141212463,
      "learning_rate": 3.806023374435663e-08,
      "loss": 0.0034,
      "num_tokens": 119060922.0,
      "reward": 0.19881045899819583,
      "reward_std": 0.4718875251710415,
      "rewards/cosine_scaled_reward": 0.19881044793874025,
      "rewards/format_reward": 0.0,
      "step": 174
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 499.9899787902832,
      "epoch": 1.7807351077313056,
      "grad_norm": 0.18940913677215576,
      "learning_rate": 3.4718563369743213e-08,
      "loss": -0.0418,
      "num_tokens": 119642241.0,
      "reward": 0.1533535374328494,
      "reward_std": 0.45827095210552216,
      "rewards/cosine_scaled_reward": 0.1533535262569785,
      "rewards/format_reward": 0.0,
      "step": 175
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 483.29467010498047,
      "epoch": 1.790874524714829,
      "grad_norm": 0.3907623291015625,
      "learning_rate": 3.15251375001192e-08,
      "loss": 0.0006,
      "num_tokens": 120215377.0,
      "reward": 0.30862145498394966,
      "reward_std": 0.44223641976714134,
      "rewards/cosine_scaled_reward": 0.308621431235224,
      "rewards/format_reward": 0.0,
      "step": 176
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 515.3270263671875,
      "epoch": 1.8010139416983524,
      "grad_norm": 0.18325906991958618,
      "learning_rate": 2.8480973601043955e-08,
      "loss": -0.0398,
      "num_tokens": 120811918.0,
      "reward": 0.24171121884137392,
      "reward_std": 0.36342719942331314,
      "rewards/cosine_scaled_reward": 0.24171120021492243,
      "rewards/format_reward": 0.0,
      "step": 177
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 542.7254676818848,
      "epoch": 1.8111533586818758,
      "grad_norm": 0.17345182597637177,
      "learning_rate": 2.558704158134023e-08,
      "loss": -0.0589,
      "num_tokens": 121438224.0,
      "reward": 0.10904384450986981,
      "reward_std": 0.4528394937515259,
      "rewards/cosine_scaled_reward": 0.10904384264722466,
      "rewards/format_reward": 0.0,
      "step": 178
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 543.6852951049805,
      "epoch": 1.8212927756653992,
      "grad_norm": 0.1887708306312561,
      "learning_rate": 2.2844263484068093e-08,
      "loss": -0.0194,
      "num_tokens": 122052366.0,
      "reward": 0.1703888149932027,
      "reward_std": 0.4111927505582571,
      "rewards/cosine_scaled_reward": 0.17038880474865437,
      "rewards/format_reward": 0.0,
      "step": 179
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 485.0948905944824,
      "epoch": 1.8314321926489225,
      "grad_norm": 0.20806105434894562,
      "learning_rate": 2.025351319275137e-08,
      "loss": -0.0092,
      "num_tokens": 122617979.0,
      "reward": 0.22041516425088048,
      "reward_std": 0.39602686278522015,
      "rewards/cosine_scaled_reward": 0.22041516285389662,
      "rewards/format_reward": 0.0,
      "step": 180
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 521.0446624755859,
      "epoch": 1.8415716096324461,
      "grad_norm": 0.1908193826675415,
      "learning_rate": 1.781561615294652e-08,
      "loss": -0.0301,
      "num_tokens": 123219299.0,
      "reward": 0.1940143120009452,
      "reward_std": 0.4348856508731842,
      "rewards/cosine_scaled_reward": 0.19401429709978402,
      "rewards/format_reward": 0.0,
      "step": 181
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 570.9743461608887,
      "epoch": 1.8517110266159695,
      "grad_norm": 0.17504295706748962,
      "learning_rate": 1.553134910924636e-08,
      "loss": -0.0429,
      "num_tokens": 123866996.0,
      "reward": 0.08305117121199146,
      "reward_std": 0.41529295966029167,
      "rewards/cosine_scaled_reward": 0.08305116754490882,
      "rewards/format_reward": 0.0,
      "step": 182
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 481.1294822692871,
      "epoch": 1.8618504435994931,
      "grad_norm": 0.20204633474349976,
      "learning_rate": 1.340143985779829e-08,
      "loss": -0.0051,
      "num_tokens": 124424848.0,
      "reward": 0.2932664602994919,
      "reward_std": 0.4205982144922018,
      "rewards/cosine_scaled_reward": 0.2932664528489113,
      "rewards/format_reward": 0.0,
      "step": 183
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 498.0591735839844,
      "epoch": 1.8719898605830165,
      "grad_norm": 0.18216805160045624,
      "learning_rate": 1.1426567014420297e-08,
      "loss": -0.0097,
      "num_tokens": 124997813.0,
      "reward": 0.27011361718177795,
      "reward_std": 0.42340338230133057,
      "rewards/cosine_scaled_reward": 0.27011359576135874,
      "rewards/format_reward": 0.0,
      "step": 184
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 513.774585723877,
      "epoch": 1.88212927756654,
      "grad_norm": 0.1906953901052475,
      "learning_rate": 9.607359798384784e-09,
      "loss": -0.0212,
      "num_tokens": 125573635.0,
      "reward": 0.15609056316316128,
      "reward_std": 0.46621261164546013,
      "rewards/cosine_scaled_reward": 0.15609055384993553,
      "rewards/format_reward": 0.0,
      "step": 185
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 511.1183280944824,
      "epoch": 1.8922686945500633,
      "grad_norm": 0.20339244604110718,
      "learning_rate": 7.944397831941951e-09,
      "loss": 0.0108,
      "num_tokens": 126178093.0,
      "reward": 0.12280734069645405,
      "reward_std": 0.4304603859782219,
      "rewards/cosine_scaled_reward": 0.12280733417719603,
      "rewards/format_reward": 0.0,
      "step": 186
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 583.9364051818848,
      "epoch": 1.9024081115335867,
      "grad_norm": 0.18367899954319,
      "learning_rate": 6.438210955644452e-09,
      "loss": -0.0444,
      "num_tokens": 126843724.0,
      "reward": 0.019926082342863083,
      "reward_std": 0.4689461179077625,
      "rewards/cosine_scaled_reward": 0.019926077569834888,
      "rewards/format_reward": 0.0,
      "step": 187
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 523.2076110839844,
      "epoch": 1.9125475285171103,
      "grad_norm": 0.17626607418060303,
      "learning_rate": 5.0892790595336575e-09,
      "loss": -0.0197,
      "num_tokens": 127449534.0,
      "reward": 0.11797729413956404,
      "reward_std": 0.497987262904644,
      "rewards/cosine_scaled_reward": 0.11797728785313666,
      "rewards/format_reward": 0.0,
      "step": 188
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 525.0011367797852,
      "epoch": 1.9226869455006337,
      "grad_norm": 0.1973433792591095,
      "learning_rate": 3.898031930240797e-09,
      "loss": -0.029,
      "num_tokens": 128044935.0,
      "reward": 0.10833488963544369,
      "reward_std": 0.4491695165634155,
      "rewards/cosine_scaled_reward": 0.1083348891697824,
      "rewards/format_reward": 0.0,
      "step": 189
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 548.4966850280762,
      "epoch": 1.9328263624841573,
      "grad_norm": 0.17003753781318665,
      "learning_rate": 2.8648491140513264e-09,
      "loss": 0.0062,
      "num_tokens": 128662836.0,
      "reward": 0.20282268431037664,
      "reward_std": 0.4673178493976593,
      "rewards/cosine_scaled_reward": 0.20282267965376377,
      "rewards/format_reward": 0.0,
      "step": 190
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 440.6149711608887,
      "epoch": 1.9429657794676807,
      "grad_norm": 0.21727675199508667,
      "learning_rate": 1.9900597959770505e-09,
      "loss": 0.0655,
      "num_tokens": 129182755.0,
      "reward": 0.40362408570945263,
      "reward_std": 0.34423849545419216,
      "rewards/cosine_scaled_reward": 0.403624065220356,
      "rewards/format_reward": 0.0,
      "step": 191
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 541.2031555175781,
      "epoch": 1.953105196451204,
      "grad_norm": 0.19199801981449127,
      "learning_rate": 1.2739426948732424e-09,
      "loss": -0.0442,
      "num_tokens": 129801641.0,
      "reward": 0.19715721579268575,
      "reward_std": 0.4420705772936344,
      "rewards/cosine_scaled_reward": 0.19715720368549228,
      "rewards/format_reward": 0.0,
      "step": 192
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 526.9609565734863,
      "epoch": 1.9632446134347274,
      "grad_norm": 0.22452791035175323,
      "learning_rate": 7.16725974635568e-10,
      "loss": 0.0124,
      "num_tokens": 130399638.0,
      "reward": 0.20976402005180717,
      "reward_std": 0.4755300283432007,
      "rewards/cosine_scaled_reward": 0.20976400654762983,
      "rewards/format_reward": 0.0,
      "step": 193
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 550.6752510070801,
      "epoch": 1.9733840304182508,
      "grad_norm": 0.18797193467617035,
      "learning_rate": 3.185871715041255e-10,
      "loss": -0.035,
      "num_tokens": 131025891.0,
      "reward": 0.1596639035269618,
      "reward_std": 0.4088666681200266,
      "rewards/cosine_scaled_reward": 0.15966389642562717,
      "rewards/format_reward": 0.0,
      "step": 194
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 447.8102798461914,
      "epoch": 1.9835234474017744,
      "grad_norm": 0.19335201382637024,
      "learning_rate": 7.96531374983589e-11,
      "loss": -0.0162,
      "num_tokens": 131547609.0,
      "reward": 0.2565318220295012,
      "reward_std": 0.46955981478095055,
      "rewards/cosine_scaled_reward": 0.25653181644156575,
      "rewards/format_reward": 0.0,
      "step": 195
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 516.494140625,
      "epoch": 1.9936628643852978,
      "grad_norm": 0.180581733584404,
      "learning_rate": 0.0,
      "loss": -0.0093,
      "num_tokens": 132131440.0,
      "reward": 0.08059131819754839,
      "reward_std": 0.5266188904643059,
      "rewards/cosine_scaled_reward": 0.08059131726622581,
      "rewards/format_reward": 0.0,
      "step": 196
    },
    {
      "epoch": 1.9936628643852978,
      "step": 196,
      "total_flos": 0.0,
      "train_loss": -0.019965038798853137,
      "train_runtime": 31425.8764,
      "train_samples_per_second": 0.703,
      "train_steps_per_second": 0.006
    }
  ],
  "logging_steps": 1,
  "max_steps": 196,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}