{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.995334370139968,
  "eval_steps": 500,
  "global_step": 450,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 766.9769439697266,
      "epoch": 0.049766718506998445,
      "grad_norm": 0.1882091611623764,
      "kl": 0.0,
      "learning_rate": 7.142857142857142e-08,
      "loss": 0.0,
      "reward": 0.056659227251657285,
      "reward_std": 0.0960957952629542,
      "rewards/equation_reward_func": 0.0030877977133059176,
      "rewards/format_reward_func": 0.05357142974389717,
      "step": 2
    },
    {
      "completion_length": 761.3757600784302,
      "epoch": 0.09953343701399689,
      "grad_norm": 0.1693430244922638,
      "kl": 3.890312857635081e-05,
      "learning_rate": 1.4285714285714285e-07,
      "loss": 0.0,
      "reward": 0.06296131052658893,
      "reward_std": 0.09832033820566721,
      "rewards/equation_reward_func": 0.0026934524921671255,
      "rewards/format_reward_func": 0.06026785844005644,
      "step": 4
    },
    {
      "completion_length": 765.7619190216064,
      "epoch": 0.14930015552099535,
      "grad_norm": 0.18038436770439148,
      "kl": 0.0003851112737720541,
      "learning_rate": 2.1428571428571426e-07,
      "loss": 0.0,
      "reward": 0.054226191830821335,
      "reward_std": 0.09056254595634528,
      "rewards/equation_reward_func": 0.003630952454841463,
      "rewards/format_reward_func": 0.05059523927047849,
      "step": 6
    },
    {
      "completion_length": 771.5922727584839,
      "epoch": 0.19906687402799378,
      "grad_norm": 0.1860150694847107,
      "kl": 0.0017208610979650985,
      "learning_rate": 2.857142857142857e-07,
      "loss": 0.0,
      "reward": 0.07269345303211594,
      "reward_std": 0.11444784759078175,
      "rewards/equation_reward_func": 0.0034970239394169766,
      "rewards/format_reward_func": 0.06919643009314314,
      "step": 8
    },
    {
      "completion_length": 768.6547737121582,
      "epoch": 0.24883359253499224,
      "grad_norm": 0.18981724977493286,
      "kl": 0.007042938013910316,
      "learning_rate": 3.5714285714285716e-07,
      "loss": 0.0,
      "reward": 0.06845982308732346,
      "reward_std": 0.11329028639011085,
      "rewards/equation_reward_func": 0.002983631022289046,
      "rewards/format_reward_func": 0.06547619227785617,
      "step": 10
    },
    {
      "completion_length": 764.2269496917725,
      "epoch": 0.2986003110419907,
      "grad_norm": 0.15985120832920074,
      "kl": 0.030209030941477977,
      "learning_rate": 4.285714285714285e-07,
      "loss": 0.0,
      "reward": 0.07950148951204028,
      "reward_std": 0.12200771895004436,
      "rewards/equation_reward_func": 0.005096726257761475,
      "rewards/format_reward_func": 0.07440476369811222,
      "step": 12
    },
    {
      "completion_length": 777.1413831710815,
      "epoch": 0.3483670295489891,
      "grad_norm": 0.3316055238246918,
      "kl": 0.24638308194698766,
      "learning_rate": 5e-07,
      "loss": 0.0002,
      "reward": 0.06816220411565155,
      "reward_std": 0.10573185497196391,
      "rewards/equation_reward_func": 0.004174107223661849,
      "rewards/format_reward_func": 0.06398809689562768,
      "step": 14
    },
    {
      "completion_length": 766.7388534545898,
      "epoch": 0.39813374805598756,
      "grad_norm": 0.22750402987003326,
      "kl": 0.36214487318648025,
      "learning_rate": 4.999740409224932e-07,
      "loss": 0.0004,
      "reward": 0.08724702637118753,
      "reward_std": 0.1364312534351484,
      "rewards/equation_reward_func": 0.003169642919601756,
      "rewards/format_reward_func": 0.08407738310052082,
      "step": 16
    },
    {
      "completion_length": 750.9323081970215,
      "epoch": 0.447900466562986,
      "grad_norm": 0.14867445826530457,
      "kl": 0.1888283666339703,
      "learning_rate": 4.998961690809627e-07,
      "loss": 0.0002,
      "reward": 0.0880877989839064,
      "reward_std": 0.13909767809673212,
      "rewards/equation_reward_func": 0.004010416825622087,
      "rewards/format_reward_func": 0.0840773832751438,
      "step": 18
    },
    {
      "completion_length": 761.3616189956665,
      "epoch": 0.4976671850699845,
      "grad_norm": 0.17096517980098724,
      "kl": 0.1737799496622756,
      "learning_rate": 4.997664006472578e-07,
      "loss": 0.0002,
      "reward": 0.09410714514524443,
      "reward_std": 0.1441169898607768,
      "rewards/equation_reward_func": 0.004077381056049489,
      "rewards/format_reward_func": 0.09002976410556585,
      "step": 20
    },
    {
      "completion_length": 749.1599817276001,
      "epoch": 0.5474339035769828,
      "grad_norm": 0.16154693067073822,
      "kl": 0.11445438460214064,
      "learning_rate": 4.995847625707292e-07,
      "loss": 0.0001,
      "reward": 0.11768601481526275,
      "reward_std": 0.1828837317698344,
      "rewards/equation_reward_func": 0.006822916879173135,
      "rewards/format_reward_func": 0.1108630983508192,
      "step": 22
    },
    {
      "completion_length": 762.6986713409424,
      "epoch": 0.5972006220839814,
      "grad_norm": 0.1822424829006195,
      "kl": 0.13439486949937418,
      "learning_rate": 4.993512925726318e-07,
      "loss": 0.0001,
      "reward": 0.12162946694297716,
      "reward_std": 0.19308959832414985,
      "rewards/equation_reward_func": 0.0055580357930011814,
      "rewards/format_reward_func": 0.11607143172295764,
      "step": 24
    },
    {
      "completion_length": 744.5833473205566,
      "epoch": 0.6469673405909798,
      "grad_norm": 0.20237824320793152,
      "kl": 0.1873470548307523,
      "learning_rate": 4.990660391382923e-07,
      "loss": 0.0002,
      "reward": 0.1205357170838397,
      "reward_std": 0.18955813159118406,
      "rewards/equation_reward_func": 0.005208333370319451,
      "rewards/format_reward_func": 0.11532738379901275,
      "step": 26
    },
    {
      "completion_length": 740.8712930679321,
      "epoch": 0.6967340590979783,
      "grad_norm": 0.20918512344360352,
      "kl": 0.29402227653190494,
      "learning_rate": 4.987290615070384e-07,
      "loss": 0.0003,
      "reward": 0.14720238508743932,
      "reward_std": 0.21827496998594142,
      "rewards/equation_reward_func": 0.006577381089300616,
      "rewards/format_reward_func": 0.14062500419095159,
      "step": 28
    },
    {
      "completion_length": 730.9189138412476,
      "epoch": 0.7465007776049767,
      "grad_norm": 0.1985033005475998,
      "kl": 0.4376440930645913,
      "learning_rate": 4.983404296598978e-07,
      "loss": 0.0004,
      "reward": 0.17537202849052846,
      "reward_std": 0.24408150045201182,
      "rewards/equation_reward_func": 0.009449404875340406,
      "rewards/format_reward_func": 0.16592262318590656,
      "step": 30
    },
    {
      "completion_length": 722.1607275009155,
      "epoch": 0.7962674961119751,
      "grad_norm": 0.23082222044467926,
      "kl": 0.5781177286989987,
      "learning_rate": 4.979002243050646e-07,
      "loss": 0.0006,
      "reward": 0.16087798058288172,
      "reward_std": 0.23381112061906606,
      "rewards/equation_reward_func": 0.005372023912059376,
      "rewards/format_reward_func": 0.15550595690729097,
      "step": 32
    },
    {
      "completion_length": 715.540937423706,
      "epoch": 0.8460342146189735,
      "grad_norm": 0.2513209283351898,
      "kl": 0.5908648972399533,
      "learning_rate": 4.974085368611381e-07,
      "loss": 0.0006,
      "reward": 0.19839285931084305,
      "reward_std": 0.2854806184768677,
      "rewards/equation_reward_func": 0.007916666922028526,
      "rewards/format_reward_func": 0.19047619437333196,
      "step": 34
    },
    {
      "completion_length": 733.3735246658325,
      "epoch": 0.895800933125972,
      "grad_norm": 0.7895224690437317,
      "kl": 0.974291667342186,
      "learning_rate": 4.968654694381379e-07,
      "loss": 0.001,
      "reward": 0.2064955406531226,
      "reward_std": 0.29307361885730643,
      "rewards/equation_reward_func": 0.01006696438525978,
      "rewards/format_reward_func": 0.19642857578583062,
      "step": 36
    },
    {
      "completion_length": 712.1808166503906,
      "epoch": 0.9455676516329704,
      "grad_norm": 0.27443817257881165,
      "kl": 0.5672950851731002,
      "learning_rate": 4.962711348162987e-07,
      "loss": 0.0006,
      "reward": 0.23159970679262187,
      "reward_std": 0.30910343912546523,
      "rewards/equation_reward_func": 0.009129464311627089,
      "rewards/format_reward_func": 0.22247024369426072,
      "step": 38
    },
    {
      "completion_length": 724.0312623977661,
      "epoch": 0.995334370139969,
      "grad_norm": 0.2718716859817505,
      "kl": 0.7035379374865443,
      "learning_rate": 4.956256564226487e-07,
      "loss": 0.0007,
      "reward": 0.22910714593308512,
      "reward_std": 0.31004760426003486,
      "rewards/equation_reward_func": 0.009613095318854903,
      "rewards/format_reward_func": 0.21949405351188034,
      "step": 40
    },
    {
      "completion_length": 689.2218210320724,
      "epoch": 1.0248833592534992,
      "grad_norm": 0.26647260785102844,
      "kl": 0.780534968172249,
      "learning_rate": 4.949291683053768e-07,
      "loss": 0.0005,
      "reward": 0.27011278850075443,
      "reward_std": 0.3508555128386146,
      "rewards/equation_reward_func": 0.010714285780283573,
      "rewards/format_reward_func": 0.2593985022486825,
      "step": 42
    },
    {
      "completion_length": 679.3750143051147,
      "epoch": 1.0746500777604977,
      "grad_norm": 0.3014063239097595,
      "kl": 0.8605983196757734,
      "learning_rate": 4.941818151059955e-07,
      "loss": 0.0009,
      "reward": 0.26439732742437627,
      "reward_std": 0.34150049267918803,
      "rewards/equation_reward_func": 0.01290922645421233,
      "rewards/format_reward_func": 0.25148810172686353,
      "step": 44
    },
    {
      "completion_length": 678.1852836608887,
      "epoch": 1.124416796267496,
      "grad_norm": 0.2869322597980499,
      "kl": 1.1382544473744929,
      "learning_rate": 4.933837520293017e-07,
      "loss": 0.0011,
      "reward": 0.28466518549248576,
      "reward_std": 0.3586017652414739,
      "rewards/equation_reward_func": 0.010855655014893273,
      "rewards/format_reward_func": 0.27380952949170023,
      "step": 46
    },
    {
      "completion_length": 665.9114689826965,
      "epoch": 1.1741835147744946,
      "grad_norm": 0.4286658465862274,
      "kl": 1.101041450863704,
      "learning_rate": 4.925351448111454e-07,
      "loss": 0.0011,
      "reward": 0.35581102105788887,
      "reward_std": 0.3954307301901281,
      "rewards/equation_reward_func": 0.012805059952370357,
      "rewards/format_reward_func": 0.34300595964305103,
      "step": 48
    },
    {
      "completion_length": 648.8006086349487,
      "epoch": 1.223950233281493,
      "grad_norm": 0.6011645793914795,
      "kl": 1.5983671792782843,
      "learning_rate": 4.91636169684011e-07,
      "loss": 0.0016,
      "reward": 0.36582590208854526,
      "reward_std": 0.40516745822969824,
      "rewards/equation_reward_func": 0.017611607605431345,
      "rewards/format_reward_func": 0.34821429441217333,
      "step": 50
    },
    {
      "completion_length": 637.3727793693542,
      "epoch": 1.2737169517884914,
      "grad_norm": 0.4506984353065491,
      "kl": 1.99405955709517,
      "learning_rate": 4.906870133404186e-07,
      "loss": 0.002,
      "reward": 0.37937500968109816,
      "reward_std": 0.40684798220172524,
      "rewards/equation_reward_func": 0.01851190527304425,
      "rewards/format_reward_func": 0.3608631034148857,
      "step": 52
    },
    {
      "completion_length": 611.6599831581116,
      "epoch": 1.32348367029549,
      "grad_norm": 0.3486514985561371,
      "kl": 1.575268767774105,
      "learning_rate": 4.896878728941531e-07,
      "loss": 0.0016,
      "reward": 0.41054316400550306,
      "reward_std": 0.402535158675164,
      "rewards/equation_reward_func": 0.014709821767610265,
      "rewards/format_reward_func": 0.39583334256894886,
      "step": 54
    },
    {
      "completion_length": 588.9256038665771,
      "epoch": 1.3732503888024883,
      "grad_norm": 0.35510003566741943,
      "kl": 1.2928779851645231,
      "learning_rate": 4.886389558393284e-07,
      "loss": 0.0013,
      "reward": 0.44056548294611275,
      "reward_std": 0.4229291075025685,
      "rewards/equation_reward_func": 0.020922619543853216,
      "rewards/format_reward_func": 0.41964286752045155,
      "step": 56
    },
    {
      "completion_length": 568.6056680679321,
      "epoch": 1.4230171073094868,
      "grad_norm": 0.4101411998271942,
      "kl": 2.0492603620514274,
      "learning_rate": 4.875404800072976e-07,
      "loss": 0.002,
      "reward": 0.4984003081917763,
      "reward_std": 0.47362194606103003,
      "rewards/equation_reward_func": 0.020721726599731483,
      "rewards/format_reward_func": 0.47767858440056443,
      "step": 58
    },
    {
      "completion_length": 566.1391491889954,
      "epoch": 1.4727838258164851,
      "grad_norm": 0.9513494968414307,
      "kl": 2.287347287638113,
      "learning_rate": 4.86392673521415e-07,
      "loss": 0.0023,
      "reward": 0.496495544211939,
      "reward_std": 0.4519720033276826,
      "rewards/equation_reward_func": 0.021049107548606116,
      "rewards/format_reward_func": 0.475446441443637,
      "step": 60
    },
    {
      "completion_length": 555.5959930419922,
      "epoch": 1.5225505443234837,
      "grad_norm": 3.8236348628997803,
      "kl": 4.3232480408623815,
      "learning_rate": 4.851957747496606e-07,
      "loss": 0.0043,
      "reward": 0.5090178686659783,
      "reward_std": 0.45009147142991424,
      "rewards/equation_reward_func": 0.02241071459138766,
      "rewards/format_reward_func": 0.48660715692676604,
      "step": 62
    },
    {
      "completion_length": 545.1212873458862,
      "epoch": 1.5723172628304822,
      "grad_norm": 1.2665306329727173,
      "kl": 2.19858290348202,
      "learning_rate": 4.839500322551386e-07,
      "loss": 0.0022,
      "reward": 0.5529836406931281,
      "reward_std": 0.44346457067877054,
      "rewards/equation_reward_func": 0.021733631443566992,
      "rewards/format_reward_func": 0.5312500153668225,
      "step": 64
    },
    {
      "completion_length": 519.7120614051819,
      "epoch": 1.6220839813374806,
      "grad_norm": 0.6547808051109314,
      "kl": 2.3480111258104444,
      "learning_rate": 4.826557047444563e-07,
      "loss": 0.0023,
      "reward": 0.576257451903075,
      "reward_std": 0.45106819295324385,
      "rewards/equation_reward_func": 0.023430060044120182,
      "rewards/format_reward_func": 0.5528273996897042,
      "step": 66
    },
    {
      "completion_length": 500.1666741371155,
      "epoch": 1.6718506998444789,
      "grad_norm": 1.3800803422927856,
      "kl": 3.120940439403057,
      "learning_rate": 4.813130610139993e-07,
      "loss": 0.0031,
      "reward": 0.5781473349779844,
      "reward_std": 0.4549011045601219,
      "rewards/equation_reward_func": 0.02531994113815017,
      "rewards/format_reward_func": 0.5528274001553655,
      "step": 68
    },
    {
      "completion_length": 500.9419732093811,
      "epoch": 1.7216174183514774,
      "grad_norm": 1.0396491289138794,
      "kl": 6.374570010229945,
      "learning_rate": 4.799223798941089e-07,
      "loss": 0.0064,
      "reward": 0.6029836446978152,
      "reward_std": 0.4363493148703128,
      "rewards/equation_reward_func": 0.02932291675824672,
      "rewards/format_reward_func": 0.5736607322469354,
      "step": 70
    },
    {
      "completion_length": 469.58408641815186,
      "epoch": 1.771384136858476,
      "grad_norm": 1.1942294836044312,
      "kl": 24.733395729679614,
      "learning_rate": 4.78483950191177e-07,
      "loss": 0.0247,
      "reward": 0.6590253091417253,
      "reward_std": 0.43813786492682993,
      "rewards/equation_reward_func": 0.03328125060033926,
      "rewards/format_reward_func": 0.6257440662011504,
      "step": 72
    },
    {
      "completion_length": 457.6049189567566,
      "epoch": 1.8211508553654743,
      "grad_norm": 1.2962743043899536,
      "kl": 5.970435416325927,
      "learning_rate": 4.769980706276687e-07,
      "loss": 0.006,
      "reward": 0.6895312615670264,
      "reward_std": 0.4353426368907094,
      "rewards/equation_reward_func": 0.0370014887466823,
      "rewards/format_reward_func": 0.6525297802872956,
      "step": 74
    },
    {
      "completion_length": 442.77977418899536,
      "epoch": 1.8709175738724728,
      "grad_norm": 6.384627819061279,
      "kl": 5289.785693533719,
      "learning_rate": 4.7546504978008595e-07,
      "loss": 5.2898,
      "reward": 0.7037648935802281,
      "reward_std": 0.41037949989549816,
      "rewards/equation_reward_func": 0.04081845350447111,
      "rewards/format_reward_func": 0.6629464458674192,
      "step": 76
    },
    {
      "completion_length": 438.72842931747437,
      "epoch": 1.9206842923794714,
      "grad_norm": 1.6061675548553467,
      "kl": 11.556056022644043,
      "learning_rate": 4.738852060148848e-07,
      "loss": 0.0116,
      "reward": 0.7011681636795402,
      "reward_std": 0.43840571492910385,
      "rewards/equation_reward_func": 0.04789434606209397,
      "rewards/format_reward_func": 0.6532738278619945,
      "step": 78
    },
    {
      "completion_length": 412.28348779678345,
      "epoch": 1.9704510108864697,
      "grad_norm": 2.0230860710144043,
      "kl": 5.730300394818187,
      "learning_rate": 4.722588674223593e-07,
      "loss": 0.0057,
      "reward": 0.7024256153963506,
      "reward_std": 0.4186383159831166,
      "rewards/equation_reward_func": 0.04840773875184823,
      "rewards/format_reward_func": 0.654017876368016,
      "step": 80
    },
    {
      "completion_length": 389.4373498213919,
      "epoch": 2.0,
      "grad_norm": 1.0441092252731323,
      "kl": 5.438737201063256,
      "learning_rate": 4.70586371748506e-07,
      "loss": 0.0032,
      "reward": 0.7210902367767534,
      "reward_std": 0.41276825533101436,
      "rewards/equation_reward_func": 0.049411029113750705,
      "rewards/format_reward_func": 0.6716792152116173,
      "step": 82
    },
    {
      "completion_length": 378.0401859283447,
      "epoch": 2.0497667185069983,
      "grad_norm": 3.568232297897339,
      "kl": 14.175606586039066,
      "learning_rate": 4.6886806632488363e-07,
      "loss": 0.0142,
      "reward": 0.7700893068686128,
      "reward_std": 0.4058987263124436,
      "rewards/equation_reward_func": 0.057291668590551126,
      "rewards/format_reward_func": 0.7127976361662149,
      "step": 84
    },
    {
      "completion_length": 365.3497085571289,
      "epoch": 2.099533437013997,
      "grad_norm": 4.238552093505859,
      "kl": 20.433330222964287,
      "learning_rate": 4.6710430799648143e-07,
      "loss": 0.0204,
      "reward": 0.7515401933342218,
      "reward_std": 0.4172304559033364,
      "rewards/equation_reward_func": 0.06031994195654988,
      "rewards/format_reward_func": 0.6912202551029623,
      "step": 86
    },
    {
      "completion_length": 357.0818510055542,
      "epoch": 2.1493001555209954,
      "grad_norm": 26.27466583251953,
      "kl": 24.954067958518863,
      "learning_rate": 4.652954630476127e-07,
      "loss": 0.025,
      "reward": 0.8057515062391758,
      "reward_std": 0.37641601357609034,
      "rewards/equation_reward_func": 0.0631919659790583,
      "rewards/format_reward_func": 0.7425595386885107,
      "step": 88
    },
    {
      "completion_length": 361.0506021976471,
      "epoch": 2.1990668740279937,
      "grad_norm": 28.50054359436035,
      "kl": 18.05073036905378,
      "learning_rate": 4.6344190712584713e-07,
      "loss": 0.0181,
      "reward": 0.8062277045100927,
      "reward_std": 0.3896879474632442,
      "rewards/equation_reward_func": 0.07259672792861238,
      "rewards/format_reward_func": 0.7336309677921236,
      "step": 90
    },
    {
      "completion_length": 351.3601257801056,
      "epoch": 2.248833592534992,
      "grad_norm": 6.563501834869385,
      "kl": 13.980788078159094,
      "learning_rate": 4.615440251639995e-07,
      "loss": 0.014,
      "reward": 0.8148214407265186,
      "reward_std": 0.3750967951491475,
      "rewards/equation_reward_func": 0.07449404880753718,
      "rewards/format_reward_func": 0.7403273959644139,
      "step": 92
    },
    {
      "completion_length": 340.1547691822052,
      "epoch": 2.298600311041991,
      "grad_norm": 4.388554573059082,
      "kl": 28.781219646334648,
      "learning_rate": 4.596022113001894e-07,
      "loss": 0.0288,
      "reward": 0.8234449606388807,
      "reward_std": 0.3968581879744306,
      "rewards/equation_reward_func": 0.0779092287120875,
      "rewards/format_reward_func": 0.7455357294529676,
      "step": 94
    },
    {
      "completion_length": 325.18229627609253,
      "epoch": 2.348367029548989,
      "grad_norm": 6.165853500366211,
      "kl": 38.15451553463936,
      "learning_rate": 4.576168687959895e-07,
      "loss": 0.0382,
      "reward": 0.8252604268491268,
      "reward_std": 0.3835659438627772,
      "rewards/equation_reward_func": 0.07674851396586746,
      "rewards/format_reward_func": 0.7485119192861021,
      "step": 96
    },
    {
      "completion_length": 325.9955415725708,
      "epoch": 2.3981337480559874,
      "grad_norm": 4.9900288581848145,
      "kl": 29.846170015633106,
      "learning_rate": 4.555884099526793e-07,
      "loss": 0.0298,
      "reward": 0.836130972020328,
      "reward_std": 0.3841063065920025,
      "rewards/equation_reward_func": 0.08017857247614302,
      "rewards/format_reward_func": 0.7559523954987526,
      "step": 98
    },
    {
      "completion_length": 325.5736680030823,
      "epoch": 2.447900466562986,
      "grad_norm": 6.271454334259033,
      "kl": 39.699058301746845,
      "learning_rate": 4.5351725602562174e-07,
      "loss": 0.0397,
      "reward": 0.8406399013474584,
      "reward_std": 0.38198388437740505,
      "rewards/equation_reward_func": 0.08022321574389935,
      "rewards/format_reward_func": 0.7604166809469461,
      "step": 100
    },
    {
      "completion_length": 306.04911255836487,
      "epoch": 2.4976671850699845,
      "grad_norm": 59.40867614746094,
      "kl": 39.07959423214197,
      "learning_rate": 4.514038371367791e-07,
      "loss": 0.0391,
      "reward": 0.8590625217184424,
      "reward_std": 0.3790609473362565,
      "rewards/equation_reward_func": 0.0897172634722665,
      "rewards/format_reward_func": 0.7693452518433332,
      "step": 102
    },
    {
      "completion_length": 305.7552146911621,
      "epoch": 2.547433903576983,
      "grad_norm": 7.068971157073975,
      "kl": 41.63126154243946,
      "learning_rate": 4.4924859218538936e-07,
      "loss": 0.0416,
      "reward": 0.8449925715103745,
      "reward_std": 0.3808569728862494,
      "rewards/equation_reward_func": 0.0853199407865759,
      "rewards/format_reward_func": 0.7596726333722472,
      "step": 104
    },
    {
      "completion_length": 304.3958387374878,
      "epoch": 2.5972006220839816,
      "grad_norm": 16.367332458496094,
      "kl": 61.27933597564697,
      "learning_rate": 4.470519687568185e-07,
      "loss": 0.0613,
      "reward": 0.8741964548826218,
      "reward_std": 0.3497261651791632,
      "rewards/equation_reward_func": 0.08625000189931598,
      "rewards/format_reward_func": 0.7879464407451451,
      "step": 106
    },
    {
      "completion_length": 290.72768449783325,
      "epoch": 2.64696734059098,
      "grad_norm": 9.564668655395508,
      "kl": 56.39366825670004,
      "learning_rate": 4.4481442302960923e-07,
      "loss": 0.0564,
      "reward": 0.8775223465636373,
      "reward_std": 0.3703669502865523,
      "rewards/equation_reward_func": 0.0947842278110329,
      "rewards/format_reward_func": 0.7827381081879139,
      "step": 108
    },
    {
      "completion_length": 285.9531292915344,
      "epoch": 2.6967340590979783,
      "grad_norm": 9.019147872924805,
      "kl": 57.990618377923965,
      "learning_rate": 4.4253641968074505e-07,
      "loss": 0.058,
      "reward": 0.8902678797021508,
      "reward_std": 0.3635682419408113,
      "rewards/equation_reward_func": 0.10083333519287407,
      "rewards/format_reward_func": 0.7894345363602042,
      "step": 110
    },
    {
      "completion_length": 268.639142036438,
      "epoch": 2.7465007776049766,
      "grad_norm": 11.662919044494629,
      "kl": 45.52927339076996,
      "learning_rate": 4.402184317891501e-07,
      "loss": 0.0455,
      "reward": 0.897909251973033,
      "reward_std": 0.33927053899969906,
      "rewards/equation_reward_func": 0.101034227933269,
      "rewards/format_reward_func": 0.7968750121071935,
      "step": 112
    },
    {
      "completion_length": 275.512654542923,
      "epoch": 2.796267496111975,
      "grad_norm": 52.24915313720703,
      "kl": 80.93976804614067,
      "learning_rate": 4.37860940737443e-07,
      "loss": 0.0809,
      "reward": 0.9052157932892442,
      "reward_std": 0.33712050085887313,
      "rewards/equation_reward_func": 0.09941220440668985,
      "rewards/format_reward_func": 0.8058035830035806,
      "step": 114
    },
    {
      "completion_length": 276.7753026485443,
      "epoch": 2.8460342146189737,
      "grad_norm": 6.988640308380127,
      "kl": 73.49725632369518,
      "learning_rate": 4.354644361119671e-07,
      "loss": 0.0735,
      "reward": 0.8936904948204756,
      "reward_std": 0.35725354170426726,
      "rewards/equation_reward_func": 0.1020238118362613,
      "rewards/format_reward_func": 0.791666679084301,
      "step": 116
    },
    {
      "completion_length": 281.043904542923,
      "epoch": 2.895800933125972,
      "grad_norm": 14.06199836730957,
      "kl": 67.49630101025105,
      "learning_rate": 4.3302941560111716e-07,
      "loss": 0.0675,
      "reward": 0.8816443579271436,
      "reward_std": 0.36942707875277847,
      "rewards/equation_reward_func": 0.10337053792318329,
      "rewards/format_reward_func": 0.7782738227397203,
      "step": 118
    },
    {
      "completion_length": 275.93155312538147,
      "epoch": 2.9455676516329703,
      "grad_norm": 7.618069648742676,
      "kl": 61.22661107778549,
      "learning_rate": 4.3055638489198236e-07,
      "loss": 0.0612,
      "reward": 0.895319958217442,
      "reward_std": 0.3638994665816426,
      "rewards/equation_reward_func": 0.1006770851672627,
      "rewards/format_reward_func": 0.7946428693830967,
      "step": 120
    },
    {
      "completion_length": 272.8802127838135,
      "epoch": 2.995334370139969,
      "grad_norm": 17.16042137145996,
      "kl": 92.38909322023392,
      "learning_rate": 4.280458575653296e-07,
      "loss": 0.0924,
      "reward": 0.9042931590229273,
      "reward_std": 0.3583536515943706,
      "rewards/equation_reward_func": 0.1156026819953695,
      "rewards/format_reward_func": 0.7886904887855053,
      "step": 122
    },
    {
      "completion_length": 267.62281196995787,
      "epoch": 3.0248833592534994,
      "grad_norm": 9.52210521697998,
      "kl": 76.53048851615505,
      "learning_rate": 4.2549835498894665e-07,
      "loss": 0.0454,
      "reward": 0.8889097963508806,
      "reward_std": 0.3777208249819906,
      "rewards/equation_reward_func": 0.09943609174929167,
      "rewards/format_reward_func": 0.7894736967588726,
      "step": 124
    },
    {
      "completion_length": 253.41592621803284,
      "epoch": 3.0746500777604977,
      "grad_norm": 12.951403617858887,
      "kl": 69.25402422249317,
      "learning_rate": 4.229144062093679e-07,
      "loss": 0.0693,
      "reward": 0.9256026921793818,
      "reward_std": 0.33047901722602546,
      "rewards/equation_reward_func": 0.11161458486458287,
      "rewards/format_reward_func": 0.8139881063252687,
      "step": 126
    },
    {
      "completion_length": 260.8653337955475,
      "epoch": 3.124416796267496,
      "grad_norm": 11.296874046325684,
      "kl": 92.7376360297203,
      "learning_rate": 4.2029454784200675e-07,
      "loss": 0.0927,
      "reward": 0.9119196608662605,
      "reward_std": 0.34998451662249863,
      "rewards/equation_reward_func": 0.10686012083897367,
      "rewards/format_reward_func": 0.8050595354288816,
      "step": 128
    },
    {
      "completion_length": 264.3511941432953,
      "epoch": 3.1741835147744943,
      "grad_norm": 13.988784790039062,
      "kl": 95.21338140964508,
      "learning_rate": 4.1763932395971433e-07,
      "loss": 0.0952,
      "reward": 0.9192336574196815,
      "reward_std": 0.3349384202156216,
      "rewards/equation_reward_func": 0.11119791801320389,
      "rewards/format_reward_func": 0.8080357257276773,
      "step": 130
    },
    {
      "completion_length": 254.3474748134613,
      "epoch": 3.223950233281493,
      "grad_norm": 10.847155570983887,
      "kl": 63.423495560884476,
      "learning_rate": 4.1494928597979117e-07,
      "loss": 0.0634,
      "reward": 0.9145759223029017,
      "reward_std": 0.33581931074149907,
      "rewards/equation_reward_func": 0.10281994252000004,
      "rewards/format_reward_func": 0.811755963601172,
      "step": 132
    },
    {
      "completion_length": 250.99033308029175,
      "epoch": 3.2737169517884914,
      "grad_norm": 18.836917877197266,
      "kl": 84.4757269024849,
      "learning_rate": 4.122249925494726e-07,
      "loss": 0.0845,
      "reward": 0.9351265020668507,
      "reward_std": 0.32899824890773743,
      "rewards/equation_reward_func": 0.111465776222758,
      "rewards/format_reward_func": 0.8236607247963548,
      "step": 134
    },
    {
      "completion_length": 259.0275332927704,
      "epoch": 3.3234836702954897,
      "grad_norm": 27.394779205322266,
      "kl": 127.20685377717018,
      "learning_rate": 4.094670094299131e-07,
      "loss": 0.1272,
      "reward": 0.9070387203246355,
      "reward_std": 0.36526231234893203,
      "rewards/equation_reward_func": 0.10867559676989913,
      "rewards/format_reward_func": 0.7983631072565913,
      "step": 136
    },
    {
      "completion_length": 249.11384463310242,
      "epoch": 3.3732503888024885,
      "grad_norm": 17.246868133544922,
      "kl": 91.59087002277374,
      "learning_rate": 4.066759093786931e-07,
      "loss": 0.0916,
      "reward": 0.9155283067375422,
      "reward_std": 0.33937901724129915,
      "rewards/equation_reward_func": 0.11270089336903766,
      "rewards/format_reward_func": 0.8028273927047849,
      "step": 138
    },
    {
      "completion_length": 253.11607432365417,
      "epoch": 3.423017107309487,
      "grad_norm": 25.47807502746582,
      "kl": 66.04804906249046,
      "learning_rate": 4.038522720308732e-07,
      "loss": 0.066,
      "reward": 0.9110640026628971,
      "reward_std": 0.35778610594570637,
      "rewards/equation_reward_func": 0.10898065764922649,
      "rewards/format_reward_func": 0.802083345130086,
      "step": 140
    },
    {
      "completion_length": 256.49181938171387,
      "epoch": 3.472783825816485,
      "grad_norm": 6.512887954711914,
      "kl": 91.01838940382004,
      "learning_rate": 4.009966837786194e-07,
      "loss": 0.091,
      "reward": 0.9077157909050584,
      "reward_std": 0.34360028826631606,
      "rewards/equation_reward_func": 0.10340029885992408,
      "rewards/format_reward_func": 0.8043154878541827,
      "step": 142
    },
    {
      "completion_length": 245.66815972328186,
      "epoch": 3.522550544323484,
      "grad_norm": 18.00103759765625,
      "kl": 110.25812494754791,
      "learning_rate": 3.981097376494259e-07,
      "loss": 0.1103,
      "reward": 0.9331994261592627,
      "reward_std": 0.33608464046847075,
      "rewards/equation_reward_func": 0.11772321787429973,
      "rewards/format_reward_func": 0.8154762014746666,
      "step": 144
    },
    {
      "completion_length": 236.0089328289032,
      "epoch": 3.5723172628304822,
      "grad_norm": 10.365816116333008,
      "kl": 78.26719461381435,
      "learning_rate": 3.951920331829592e-07,
      "loss": 0.0783,
      "reward": 0.9498809697106481,
      "reward_std": 0.3238170698750764,
      "rewards/equation_reward_func": 0.11059524107258767,
      "rewards/format_reward_func": 0.8392857238650322,
      "step": 146
    },
    {
      "completion_length": 249.16890382766724,
      "epoch": 3.6220839813374806,
      "grad_norm": 19.683488845825195,
      "kl": 104.068582624197,
      "learning_rate": 3.922441763065506e-07,
      "loss": 0.1041,
      "reward": 0.9355059750378132,
      "reward_std": 0.3465363304130733,
      "rewards/equation_reward_func": 0.11556547920918092,
      "rewards/format_reward_func": 0.8199404869228601,
      "step": 148
    },
    {
      "completion_length": 236.11310124397278,
      "epoch": 3.671850699844479,
      "grad_norm": 7.490208148956299,
      "kl": 107.21742415428162,
      "learning_rate": 3.8926677920936093e-07,
      "loss": 0.1072,
      "reward": 0.9279613336548209,
      "reward_std": 0.3398581095971167,
      "rewards/equation_reward_func": 0.10950893117114902,
      "rewards/format_reward_func": 0.8184523917734623,
      "step": 150
    },
    {
      "completion_length": 232.6324450969696,
      "epoch": 3.721617418351477,
      "grad_norm": 8.263045310974121,
      "kl": 95.21342560648918,
      "learning_rate": 3.862604602152464e-07,
      "loss": 0.0952,
      "reward": 0.9363021049648523,
      "reward_std": 0.32941144425421953,
      "rewards/equation_reward_func": 0.1141294669941999,
      "rewards/format_reward_func": 0.8221726296469569,
      "step": 152
    },
    {
      "completion_length": 236.56548023223877,
      "epoch": 3.771384136858476,
      "grad_norm": 7.246555805206299,
      "kl": 99.01788094639778,
      "learning_rate": 3.8322584365434934e-07,
      "loss": 0.099,
      "reward": 0.9559821607545018,
      "reward_std": 0.33185221953317523,
      "rewards/equation_reward_func": 0.1219047627528198,
      "rewards/format_reward_func": 0.8340773908421397,
      "step": 154
    },
    {
      "completion_length": 228.0424153804779,
      "epoch": 3.8211508553654743,
      "grad_norm": 10.294706344604492,
      "kl": 93.09150391817093,
      "learning_rate": 3.8016355973344173e-07,
      "loss": 0.0931,
      "reward": 0.9507738389074802,
      "reward_std": 0.3249152133939788,
      "rewards/equation_reward_func": 0.1226488123065792,
      "rewards/format_reward_func": 0.8281250102445483,
      "step": 156
    },
    {
      "completion_length": 233.0967297554016,
      "epoch": 3.8709175738724726,
      "grad_norm": 139.66383361816406,
      "kl": 165.564805701375,
      "learning_rate": 3.7707424440504863e-07,
      "loss": 0.1656,
      "reward": 0.9361756034195423,
      "reward_std": 0.3188853694591671,
      "rewards/equation_reward_func": 0.11028273997362703,
      "rewards/format_reward_func": 0.8258928675204515,
      "step": 158
    },
    {
      "completion_length": 228.4992606639862,
      "epoch": 3.9206842923794714,
      "grad_norm": 10.041633605957031,
      "kl": 112.89981982111931,
      "learning_rate": 3.739585392353787e-07,
      "loss": 0.1129,
      "reward": 0.9519345378503203,
      "reward_std": 0.3280767105752602,
      "rewards/equation_reward_func": 0.12306547746993601,
      "rewards/format_reward_func": 0.8288690578192472,
      "step": 160
    },
    {
      "completion_length": 235.31845617294312,
      "epoch": 3.9704510108864697,
      "grad_norm": 32.37154006958008,
      "kl": 111.56759664416313,
      "learning_rate": 3.7081709127108767e-07,
      "loss": 0.1116,
      "reward": 0.9412202695384622,
      "reward_std": 0.3349445411004126,
      "rewards/equation_reward_func": 0.11681547819171101,
      "rewards/format_reward_func": 0.8244047723710537,
      "step": 162
    },
    {
      "completion_length": 229.56516787880346,
      "epoch": 4.0,
      "grad_norm": 2.2472083568573,
      "kl": 72.366690811358,
      "learning_rate": 3.6765055290490513e-07,
      "loss": 0.043,
      "reward": 0.9589724556395882,
      "reward_std": 0.32855926924630213,
      "rewards/equation_reward_func": 0.11937343593882888,
      "rewards/format_reward_func": 0.8395990070543791,
      "step": 164
    },
    {
      "completion_length": 234.28199791908264,
      "epoch": 4.049766718506999,
      "grad_norm": 5.950051784515381,
      "kl": 91.56283241510391,
      "learning_rate": 3.644595817401501e-07,
      "loss": 0.0916,
      "reward": 0.9548512073233724,
      "reward_std": 0.3294029913377017,
      "rewards/equation_reward_func": 0.12226190732326359,
      "rewards/format_reward_func": 0.8325892956927419,
      "step": 166
    },
    {
      "completion_length": 224.3154797554016,
      "epoch": 4.099533437013997,
      "grad_norm": 5.77777624130249,
      "kl": 109.78453694283962,
      "learning_rate": 3.6124484045416483e-07,
      "loss": 0.1098,
      "reward": 0.9643006175756454,
      "reward_std": 0.3065133059863001,
      "rewards/equation_reward_func": 0.11683035839814693,
      "rewards/format_reward_func": 0.8474702471867204,
      "step": 168
    },
    {
      "completion_length": 232.4531307220459,
      "epoch": 4.149300155520995,
      "grad_norm": 11.240461349487305,
      "kl": 93.93413895368576,
      "learning_rate": 3.580069966606949e-07,
      "loss": 0.0939,
      "reward": 0.9517262112349272,
      "reward_std": 0.33299871208146214,
      "rewards/equation_reward_func": 0.12285714561585337,
      "rewards/format_reward_func": 0.8288690578192472,
      "step": 170
    },
    {
      "completion_length": 238.5349748134613,
      "epoch": 4.199066874027994,
      "grad_norm": 9.954313278198242,
      "kl": 92.95667415857315,
      "learning_rate": 3.547467227712444e-07,
      "loss": 0.093,
      "reward": 0.9439657982438803,
      "reward_std": 0.32343020231928676,
      "rewards/equation_reward_func": 0.12328125181375071,
      "rewards/format_reward_func": 0.8206845344975591,
      "step": 172
    },
    {
      "completion_length": 233.2410750389099,
      "epoch": 4.248833592534992,
      "grad_norm": 8.026561737060547,
      "kl": 106.31019732356071,
      "learning_rate": 3.5146469585543386e-07,
      "loss": 0.1063,
      "reward": 0.9394568698480725,
      "reward_std": 0.33057266171090305,
      "rewards/equation_reward_func": 0.1202604187419638,
      "rewards/format_reward_func": 0.8191964393481612,
      "step": 174
    },
    {
      "completion_length": 229.54911065101624,
      "epoch": 4.298600311041991,
      "grad_norm": 15.306875228881836,
      "kl": 105.12126851081848,
      "learning_rate": 3.481615975003922e-07,
      "loss": 0.1051,
      "reward": 0.9428199585527182,
      "reward_std": 0.32898445217870176,
      "rewards/equation_reward_func": 0.11618303920840845,
      "rewards/format_reward_func": 0.8266369150951505,
      "step": 176
    },
    {
      "completion_length": 224.49479627609253,
      "epoch": 4.348367029548989,
      "grad_norm": 19.5836238861084,
      "kl": 118.43205219507217,
      "learning_rate": 3.448381136692089e-07,
      "loss": 0.1184,
      "reward": 0.9482961613684893,
      "reward_std": 0.334218667820096,
      "rewards/equation_reward_func": 0.12240327592007816,
      "rewards/format_reward_func": 0.8258928675204515,
      "step": 178
    },
    {
      "completion_length": 227.92634320259094,
      "epoch": 4.3981337480559874,
      "grad_norm": 11.565065383911133,
      "kl": 99.0534802004695,
      "learning_rate": 3.4149493455847897e-07,
      "loss": 0.0991,
      "reward": 0.9628125233575702,
      "reward_std": 0.31036914128344506,
      "rewards/equation_reward_func": 0.12055059859994799,
      "rewards/format_reward_func": 0.8422619141638279,
      "step": 180
    },
    {
      "completion_length": 228.84821903705597,
      "epoch": 4.447900466562986,
      "grad_norm": 11.746493339538574,
      "kl": 94.81702630221844,
      "learning_rate": 3.3813275445496766e-07,
      "loss": 0.0948,
      "reward": 0.9356473386287689,
      "reward_std": 0.34875215077772737,
      "rewards/equation_reward_func": 0.1201711333123967,
      "rewards/format_reward_func": 0.8154762014746666,
      "step": 182
    },
    {
      "completion_length": 230.1302137374878,
      "epoch": 4.497667185069984,
      "grad_norm": 6.934517860412598,
      "kl": 104.89102700352669,
      "learning_rate": 3.347522715914262e-07,
      "loss": 0.1049,
      "reward": 0.941220261156559,
      "reward_std": 0.3412000882672146,
      "rewards/equation_reward_func": 0.11607142945285887,
      "rewards/format_reward_func": 0.8251488199457526,
      "step": 184
    },
    {
      "completion_length": 237.3623547554016,
      "epoch": 4.547433903576983,
      "grad_norm": 8.869417190551758,
      "kl": 115.99200537800789,
      "learning_rate": 3.313541880015877e-07,
      "loss": 0.116,
      "reward": 0.9583631167188287,
      "reward_std": 0.3228510069893673,
      "rewards/equation_reward_func": 0.11982142983470112,
      "rewards/format_reward_func": 0.8385416762903333,
      "step": 186
    },
    {
      "completion_length": 214.82440757751465,
      "epoch": 4.597200622083982,
      "grad_norm": 6.815430164337158,
      "kl": 199.46320822834969,
      "learning_rate": 3.279392093743747e-07,
      "loss": 0.1995,
      "reward": 0.943742586299777,
      "reward_std": 0.3288773400709033,
      "rewards/equation_reward_func": 0.12305803812341765,
      "rewards/format_reward_func": 0.8206845344975591,
      "step": 188
    },
    {
      "completion_length": 223.35640358924866,
      "epoch": 4.6469673405909795,
      "grad_norm": 14.837751388549805,
      "kl": 92.87921611964703,
      "learning_rate": 3.245080449073459e-07,
      "loss": 0.0929,
      "reward": 0.9597470490261912,
      "reward_std": 0.3248986240942031,
      "rewards/equation_reward_func": 0.1286458355607465,
      "rewards/format_reward_func": 0.831101200543344,
      "step": 190
    },
    {
      "completion_length": 227.95610523223877,
      "epoch": 4.696734059097978,
      "grad_norm": 7.916383266448975,
      "kl": 84.76704713702202,
      "learning_rate": 3.210614071594162e-07,
      "loss": 0.0848,
      "reward": 0.9576339479535818,
      "reward_std": 0.32447043631691486,
      "rewards/equation_reward_func": 0.12281250266823918,
      "rewards/format_reward_func": 0.8348214384168386,
      "step": 192
    },
    {
      "completion_length": 230.10714721679688,
      "epoch": 4.746500777604977,
      "grad_norm": 57.35398864746094,
      "kl": 135.87937945127487,
      "learning_rate": 3.1760001190287695e-07,
      "loss": 0.1359,
      "reward": 0.9570536017417908,
      "reward_std": 0.3135519213974476,
      "rewards/equation_reward_func": 0.1185119078727439,
      "rewards/format_reward_func": 0.8385416762903333,
      "step": 194
    },
    {
      "completion_length": 223.31771206855774,
      "epoch": 4.796267496111975,
      "grad_norm": 62.215885162353516,
      "kl": 186.61959165334702,
      "learning_rate": 3.141245779747502e-07,
      "loss": 0.1866,
      "reward": 0.9525744337588549,
      "reward_std": 0.32037264411337674,
      "rewards/equation_reward_func": 0.1207291692844592,
      "rewards/format_reward_func": 0.831845248118043,
      "step": 196
    },
    {
      "completion_length": 228.9754502773285,
      "epoch": 4.846034214618974,
      "grad_norm": 16.842937469482422,
      "kl": 158.74550223350525,
      "learning_rate": 3.106358271275056e-07,
      "loss": 0.1587,
      "reward": 0.9604613315314054,
      "reward_std": 0.3215845621889457,
      "rewards/equation_reward_func": 0.13010416802717373,
      "rewards/format_reward_func": 0.8303571529686451,
      "step": 198
    },
    {
      "completion_length": 221.91443800926208,
      "epoch": 4.895800933125972,
      "grad_norm": 9.416200637817383,
      "kl": 88.87509013712406,
      "learning_rate": 3.0713448387917227e-07,
      "loss": 0.0889,
      "reward": 0.9541145944967866,
      "reward_std": 0.3186596838058904,
      "rewards/equation_reward_func": 0.12078125210246071,
      "rewards/format_reward_func": 0.8333333432674408,
      "step": 200
    },
    {
      "completion_length": 226.09821820259094,
      "epoch": 4.94556765163297,
      "grad_norm": 7.521836280822754,
      "kl": 72.75468015670776,
      "learning_rate": 3.0362127536287636e-07,
      "loss": 0.0728,
      "reward": 0.9630208564922214,
      "reward_std": 0.3088786149164662,
      "rewards/equation_reward_func": 0.12224702612729743,
      "rewards/format_reward_func": 0.84077381901443,
      "step": 202
    },
    {
      "completion_length": 220.4330394268036,
      "epoch": 4.995334370139969,
      "grad_norm": 6.782564163208008,
      "kl": 95.6797014772892,
      "learning_rate": 3.0009693117583523e-07,
      "loss": 0.0957,
      "reward": 0.9688095478340983,
      "reward_std": 0.3198074179235846,
      "rewards/equation_reward_func": 0.1258035731734708,
      "rewards/format_reward_func": 0.8430059617385268,
      "step": 204
    },
    {
      "completion_length": 223.10902605558695,
      "epoch": 5.024883359253499,
      "grad_norm": 14.31734561920166,
      "kl": 147.89477518985146,
      "learning_rate": 2.965621832278401e-07,
      "loss": 0.0878,
      "reward": 0.9550376173697019,
      "reward_std": 0.32527210092858266,
      "rewards/equation_reward_func": 0.12421052804902981,
      "rewards/format_reward_func": 0.8308270777526655,
      "step": 206
    },
    {
      "completion_length": 219.39360547065735,
      "epoch": 5.074650077760498,
      "grad_norm": 13.403369903564453,
      "kl": 121.50938338041306,
      "learning_rate": 2.9301776558925875e-07,
      "loss": 0.1215,
      "reward": 0.9764509275555611,
      "reward_std": 0.3119593682931736,
      "rewards/equation_reward_func": 0.12898065720219165,
      "rewards/format_reward_func": 0.8474702471867204,
      "step": 208
    },
    {
      "completion_length": 221.6532769203186,
      "epoch": 5.1244167962674965,
      "grad_norm": 8.643827438354492,
      "kl": 100.20702999830246,
      "learning_rate": 2.894644143385885e-07,
      "loss": 0.1002,
      "reward": 0.9445982417091727,
      "reward_std": 0.33305030199699104,
      "rewards/equation_reward_func": 0.12614583619870245,
      "rewards/format_reward_func": 0.8184523917734623,
      "step": 210
    },
    {
      "completion_length": 222.4598252773285,
      "epoch": 5.174183514774494,
      "grad_norm": 9.457352638244629,
      "kl": 77.77037671208382,
      "learning_rate": 2.859028674095937e-07,
      "loss": 0.0778,
      "reward": 0.9626786010339856,
      "reward_std": 0.32365662849042565,
      "rewards/equation_reward_func": 0.12711309758014977,
      "rewards/format_reward_func": 0.8355654859915376,
      "step": 212
    },
    {
      "completion_length": 216.18601536750793,
      "epoch": 5.223950233281493,
      "grad_norm": 5.520247459411621,
      "kl": 108.70252940058708,
      "learning_rate": 2.823338644380566e-07,
      "loss": 0.1087,
      "reward": 0.9588839625939727,
      "reward_std": 0.31889201514422894,
      "rewards/equation_reward_func": 0.12629464478231966,
      "rewards/format_reward_func": 0.8325892956927419,
      "step": 214
    },
    {
      "completion_length": 217.8199441432953,
      "epoch": 5.273716951788492,
      "grad_norm": 8.16267204284668,
      "kl": 144.65436732769012,
      "learning_rate": 2.7875814660817504e-07,
      "loss": 0.1447,
      "reward": 0.9597247261554003,
      "reward_std": 0.33260277623776346,
      "rewards/equation_reward_func": 0.12490327673731372,
      "rewards/format_reward_func": 0.8348214384168386,
      "step": 216
    },
    {
      "completion_length": 216.7961344718933,
      "epoch": 5.32348367029549,
      "grad_norm": 5.935340404510498,
      "kl": 100.77805884182453,
      "learning_rate": 2.751764564986396e-07,
      "loss": 0.1008,
      "reward": 0.9963318752124906,
      "reward_std": 0.29392986895982176,
      "rewards/equation_reward_func": 0.13174851471558213,
      "rewards/format_reward_func": 0.8645833414047956,
      "step": 218
    },
    {
      "completion_length": 220.86533164978027,
      "epoch": 5.3732503888024885,
      "grad_norm": 4.818342208862305,
      "kl": 93.73328730463982,
      "learning_rate": 2.715895379284194e-07,
      "loss": 0.0937,
      "reward": 0.9783780043944716,
      "reward_std": 0.30780787393450737,
      "rewards/equation_reward_func": 0.12941964401397854,
      "rewards/format_reward_func": 0.8489583423361182,
      "step": 220
    },
    {
      "completion_length": 213.26414012908936,
      "epoch": 5.423017107309486,
      "grad_norm": 11.288951873779297,
      "kl": 102.97140884399414,
      "learning_rate": 2.6799813580229174e-07,
      "loss": 0.103,
      "reward": 0.9846949661150575,
      "reward_std": 0.3030894915573299,
      "rewards/equation_reward_func": 0.13573660928523168,
      "rewards/format_reward_func": 0.8489583423361182,
      "step": 222
    },
    {
      "completion_length": 213.67411065101624,
      "epoch": 5.472783825816485,
      "grad_norm": 5.07380485534668,
      "kl": 106.15928809344769,
      "learning_rate": 2.6440299595614606e-07,
      "loss": 0.1062,
      "reward": 0.9686905024573207,
      "reward_std": 0.3141279687406495,
      "rewards/equation_reward_func": 0.12791666924022138,
      "rewards/format_reward_func": 0.84077381901443,
      "step": 224
    },
    {
      "completion_length": 217.79241490364075,
      "epoch": 5.522550544323484,
      "grad_norm": 13.45128345489502,
      "kl": 100.54976835846901,
      "learning_rate": 2.6080486500209347e-07,
      "loss": 0.1006,
      "reward": 0.9762128228321671,
      "reward_std": 0.30432219384238124,
      "rewards/equation_reward_func": 0.11906994262244552,
      "rewards/format_reward_func": 0.8571428656578064,
      "step": 226
    },
    {
      "completion_length": 213.45833671092987,
      "epoch": 5.572317262830482,
      "grad_norm": 5.673408031463623,
      "kl": 107.341905772686,
      "learning_rate": 2.572044901734166e-07,
      "loss": 0.1073,
      "reward": 0.987425615079701,
      "reward_std": 0.3038695892319083,
      "rewards/equation_reward_func": 0.13400297972839326,
      "rewards/format_reward_func": 0.8534226277843118,
      "step": 228
    },
    {
      "completion_length": 215.05655169487,
      "epoch": 5.6220839813374806,
      "grad_norm": 10.91450023651123,
      "kl": 104.37228363752365,
      "learning_rate": 2.536026191693893e-07,
      "loss": 0.1044,
      "reward": 0.9818452587351203,
      "reward_std": 0.3116321135312319,
      "rewards/equation_reward_func": 0.13511905062478036,
      "rewards/format_reward_func": 0.8467261996120214,
      "step": 230
    },
    {
      "completion_length": 221.6837841272354,
      "epoch": 5.671850699844479,
      "grad_norm": 7.277580738067627,
      "kl": 102.12091478705406,
      "learning_rate": 2.5e-07,
      "loss": 0.1021,
      "reward": 0.9730580523610115,
      "reward_std": 0.31462284130975604,
      "rewards/equation_reward_func": 0.13079613307490945,
      "rewards/format_reward_func": 0.8422619141638279,
      "step": 232
    },
    {
      "completion_length": 213.37202835083008,
      "epoch": 5.721617418351477,
      "grad_norm": 6.820380210876465,
      "kl": 110.825957685709,
      "learning_rate": 2.4639738083061073e-07,
      "loss": 0.1108,
      "reward": 0.9786160914227366,
      "reward_std": 0.2993557241279632,
      "rewards/equation_reward_func": 0.12370535923400894,
      "rewards/format_reward_func": 0.8549107229337096,
      "step": 234
    },
    {
      "completion_length": 219.30580830574036,
      "epoch": 5.771384136858476,
      "grad_norm": 5.3565287590026855,
      "kl": 112.17789036035538,
      "learning_rate": 2.4279550982658345e-07,
      "loss": 0.1122,
      "reward": 0.9671205542981625,
      "reward_std": 0.31208174559287727,
      "rewards/equation_reward_func": 0.12783482612576336,
      "rewards/format_reward_func": 0.8392857238650322,
      "step": 236
    },
    {
      "completion_length": 224.95908081531525,
      "epoch": 5.821150855365475,
      "grad_norm": 8.847481727600098,
      "kl": 89.92903319001198,
      "learning_rate": 2.3919513499790646e-07,
      "loss": 0.0899,
      "reward": 0.9810863379389048,
      "reward_std": 0.2922416543588042,
      "rewards/equation_reward_func": 0.12840774038340896,
      "rewards/format_reward_func": 0.8526785802096128,
      "step": 238
    },
    {
      "completion_length": 213.0193498134613,
      "epoch": 5.870917573872473,
      "grad_norm": 8.37535285949707,
      "kl": 91.88962814211845,
      "learning_rate": 2.3559700404385394e-07,
      "loss": 0.0919,
      "reward": 0.9962723441421986,
      "reward_std": 0.2967066681012511,
      "rewards/equation_reward_func": 0.133921132190153,
      "rewards/format_reward_func": 0.8623511986806989,
      "step": 240
    },
    {
      "completion_length": 215.90848636627197,
      "epoch": 5.920684292379471,
      "grad_norm": 11.996051788330078,
      "kl": 123.01588797569275,
      "learning_rate": 2.3200186419770823e-07,
      "loss": 0.123,
      "reward": 0.9946056772023439,
      "reward_std": 0.2892517475411296,
      "rewards/equation_reward_func": 0.137462800135836,
      "rewards/format_reward_func": 0.8571428656578064,
      "step": 242
    },
    {
      "completion_length": 207.04836869239807,
      "epoch": 5.970451010886469,
      "grad_norm": 20.457063674926758,
      "kl": 106.15135931968689,
      "learning_rate": 2.284104620715807e-07,
      "loss": 0.1062,
      "reward": 0.9806399028748274,
      "reward_std": 0.3068936343770474,
      "rewards/equation_reward_func": 0.12870536016998813,
      "rewards/format_reward_func": 0.8519345326349139,
      "step": 244
    },
    {
      "completion_length": 204.9586510909231,
      "epoch": 6.0,
      "grad_norm": 2.4967029094696045,
      "kl": 93.90482782062732,
      "learning_rate": 2.2482354350136043e-07,
      "loss": 0.0558,
      "reward": 0.9888596707268765,
      "reward_std": 0.30840789526700974,
      "rewards/equation_reward_func": 0.12795739620923996,
      "rewards/format_reward_func": 0.8609022639299694,
      "step": 246
    },
    {
      "completion_length": 216.3809564113617,
      "epoch": 6.049766718506999,
      "grad_norm": 14.058419227600098,
      "kl": 95.39833578467369,
      "learning_rate": 2.2124185339182496e-07,
      "loss": 0.0954,
      "reward": 0.9900744194164872,
      "reward_std": 0.2951568606076762,
      "rewards/equation_reward_func": 0.12772321642842144,
      "rewards/format_reward_func": 0.8623511986806989,
      "step": 248
    },
    {
      "completion_length": 215.97768187522888,
      "epoch": 6.099533437013997,
      "grad_norm": 8.967341423034668,
      "kl": 130.6636378467083,
      "learning_rate": 2.1766613556194344e-07,
      "loss": 0.1307,
      "reward": 0.9844643184915185,
      "reward_std": 0.3151676075067371,
      "rewards/equation_reward_func": 0.13252976315561682,
      "rewards/format_reward_func": 0.8519345326349139,
      "step": 250
    },
    {
      "completion_length": 215.37574863433838,
      "epoch": 6.149300155520995,
      "grad_norm": 8.772759437561035,
      "kl": 123.94878220558167,
      "learning_rate": 2.1409713259040628e-07,
      "loss": 0.1239,
      "reward": 0.9948214497417212,
      "reward_std": 0.2969008543295786,
      "rewards/equation_reward_func": 0.1324702415149659,
      "rewards/format_reward_func": 0.8623511986806989,
      "step": 252
    },
    {
      "completion_length": 204.1808078289032,
      "epoch": 6.199066874027994,
      "grad_norm": 8.49015998840332,
      "kl": 115.75950698554516,
      "learning_rate": 2.105355856614115e-07,
      "loss": 0.1158,
      "reward": 0.9843080537393689,
      "reward_std": 0.2868085833033547,
      "rewards/equation_reward_func": 0.12716518191155046,
      "rewards/format_reward_func": 0.8571428656578064,
      "step": 254
    },
    {
      "completion_length": 227.52753329277039,
      "epoch": 6.248833592534992,
      "grad_norm": 7.344859600067139,
      "kl": 94.37721315026283,
      "learning_rate": 2.069822344107413e-07,
      "loss": 0.0944,
      "reward": 0.9889434743672609,
      "reward_std": 0.2974336859770119,
      "rewards/equation_reward_func": 0.1258482167031616,
      "rewards/format_reward_func": 0.8630952462553978,
      "step": 256
    },
    {
      "completion_length": 213.5558066368103,
      "epoch": 6.298600311041991,
      "grad_norm": 7.32083797454834,
      "kl": 90.64998903870583,
      "learning_rate": 2.034378167721599e-07,
      "loss": 0.0907,
      "reward": 0.9939955649897456,
      "reward_std": 0.2892067184438929,
      "rewards/equation_reward_func": 0.13536458619637415,
      "rewards/format_reward_func": 0.8586309608072042,
      "step": 258
    },
    {
      "completion_length": 223.00298142433167,
      "epoch": 6.348367029548989,
      "grad_norm": 15.500726699829102,
      "kl": 107.22374087572098,
      "learning_rate": 1.9990306882416485e-07,
      "loss": 0.1072,
      "reward": 0.9919122280552983,
      "reward_std": 0.291991472709924,
      "rewards/equation_reward_func": 0.12732887221500278,
      "rewards/format_reward_func": 0.8645833414047956,
      "step": 260
    },
    {
      "completion_length": 213.41815984249115,
      "epoch": 6.3981337480559874,
      "grad_norm": 21.084169387817383,
      "kl": 135.0164725780487,
      "learning_rate": 1.9637872463712362e-07,
      "loss": 0.135,
      "reward": 0.9947991305962205,
      "reward_std": 0.2926121799973771,
      "rewards/equation_reward_func": 0.1294717276468873,
      "rewards/format_reward_func": 0.8653273889794946,
      "step": 262
    },
    {
      "completion_length": 218.77678954601288,
      "epoch": 6.447900466562986,
      "grad_norm": 7.433415412902832,
      "kl": 126.0510938167572,
      "learning_rate": 1.9286551612082773e-07,
      "loss": 0.1261,
      "reward": 0.9961086511611938,
      "reward_std": 0.29620418848935515,
      "rewards/equation_reward_func": 0.13152529881335795,
      "rewards/format_reward_func": 0.8645833414047956,
      "step": 264
    },
    {
      "completion_length": 224.87351608276367,
      "epoch": 6.497667185069984,
      "grad_norm": 10.848310470581055,
      "kl": 108.13037157058716,
      "learning_rate": 1.8936417287249446e-07,
      "loss": 0.1081,
      "reward": 0.9826934821903706,
      "reward_std": 0.29942381975706667,
      "rewards/equation_reward_func": 0.12703869334654883,
      "rewards/format_reward_func": 0.8556547705084085,
      "step": 266
    },
    {
      "completion_length": 205.85045075416565,
      "epoch": 6.547433903576983,
      "grad_norm": 5.818633556365967,
      "kl": 93.97169226408005,
      "learning_rate": 1.8587542202524985e-07,
      "loss": 0.094,
      "reward": 0.9770759316161275,
      "reward_std": 0.3137034277897328,
      "rewards/equation_reward_func": 0.1348139907931909,
      "rewards/format_reward_func": 0.8422619141638279,
      "step": 268
    },
    {
      "completion_length": 225.59821772575378,
      "epoch": 6.597200622083982,
      "grad_norm": 8.91329574584961,
      "kl": 92.77950745820999,
      "learning_rate": 1.82399988097123e-07,
      "loss": 0.0928,
      "reward": 0.9724628226831555,
      "reward_std": 0.31350986543111503,
      "rewards/equation_reward_func": 0.12722470529843122,
      "rewards/format_reward_func": 0.8452381044626236,
      "step": 270
    },
    {
      "completion_length": 210.45536065101624,
      "epoch": 6.6469673405909795,
      "grad_norm": 66.8660659790039,
      "kl": 119.46630339324474,
      "learning_rate": 1.7893859284058378e-07,
      "loss": 0.1195,
      "reward": 0.9536458635702729,
      "reward_std": 0.33827802538871765,
      "rewards/equation_reward_func": 0.12626488378737122,
      "rewards/format_reward_func": 0.8273809626698494,
      "step": 272
    },
    {
      "completion_length": 213.05655217170715,
      "epoch": 6.696734059097978,
      "grad_norm": 10.287453651428223,
      "kl": 129.00853845477104,
      "learning_rate": 1.7549195509265407e-07,
      "loss": 0.129,
      "reward": 1.006264909170568,
      "reward_std": 0.2873664147919044,
      "rewards/equation_reward_func": 0.1349851224804297,
      "rewards/format_reward_func": 0.871279769577086,
      "step": 274
    },
    {
      "completion_length": 208.0654811859131,
      "epoch": 6.746500777604977,
      "grad_norm": 9.000530242919922,
      "kl": 123.86803221702576,
      "learning_rate": 1.7206079062562536e-07,
      "loss": 0.1239,
      "reward": 1.0023512253537774,
      "reward_std": 0.2862590797012672,
      "rewards/equation_reward_func": 0.13404762104619294,
      "rewards/format_reward_func": 0.8683035792782903,
      "step": 276
    },
    {
      "completion_length": 215.65551018714905,
      "epoch": 6.796267496111975,
      "grad_norm": 7.632518768310547,
      "kl": 110.37519046664238,
      "learning_rate": 1.6864581199841226e-07,
      "loss": 0.1104,
      "reward": 0.9910863311961293,
      "reward_std": 0.3006105379899964,
      "rewards/equation_reward_func": 0.1354315506760031,
      "rewards/format_reward_func": 0.8556547705084085,
      "step": 278
    },
    {
      "completion_length": 218.75967693328857,
      "epoch": 6.846034214618974,
      "grad_norm": 8.187287330627441,
      "kl": 96.99012359976768,
      "learning_rate": 1.6524772840857388e-07,
      "loss": 0.097,
      "reward": 0.9843527041375637,
      "reward_std": 0.3118166974745691,
      "rewards/equation_reward_func": 0.13241815741639584,
      "rewards/format_reward_func": 0.8519345326349139,
      "step": 280
    },
    {
      "completion_length": 215.47321796417236,
      "epoch": 6.895800933125972,
      "grad_norm": 6.7685017585754395,
      "kl": 102.11789983510971,
      "learning_rate": 1.6186724554503237e-07,
      "loss": 0.1021,
      "reward": 0.9951934739947319,
      "reward_std": 0.29620578815229237,
      "rewards/equation_reward_func": 0.13805059966398403,
      "rewards/format_reward_func": 0.8571428656578064,
      "step": 282
    },
    {
      "completion_length": 217.3355691432953,
      "epoch": 6.94556765163297,
      "grad_norm": 7.022984504699707,
      "kl": 123.25619751214981,
      "learning_rate": 1.5850506544152103e-07,
      "loss": 0.1233,
      "reward": 0.9815848367288709,
      "reward_std": 0.3131761790718883,
      "rewards/equation_reward_func": 0.13560268166474998,
      "rewards/format_reward_func": 0.8459821520373225,
      "step": 284
    },
    {
      "completion_length": 226.38690876960754,
      "epoch": 6.995334370139969,
      "grad_norm": 7.854581832885742,
      "kl": 118.38798774778843,
      "learning_rate": 1.5516188633079107e-07,
      "loss": 0.1184,
      "reward": 0.9945535939186811,
      "reward_std": 0.2934186675120145,
      "rewards/equation_reward_func": 0.13145833683665842,
      "rewards/format_reward_func": 0.8630952462553978,
      "step": 286
    },
    {
      "completion_length": 209.86215812281557,
      "epoch": 7.024883359253499,
      "grad_norm": 7.611196994781494,
      "kl": 103.30944246994822,
      "learning_rate": 1.5183840249960784e-07,
      "loss": 0.0613,
      "reward": 1.0007143130427913,
      "reward_std": 0.2819862887263298,
      "rewards/equation_reward_func": 0.13354636826797536,
      "rewards/format_reward_func": 0.8671679277169076,
      "step": 288
    },
    {
      "completion_length": 215.6183067560196,
      "epoch": 7.074650077760498,
      "grad_norm": 189.0833740234375,
      "kl": 133.04253885149956,
      "learning_rate": 1.4853530414456612e-07,
      "loss": 0.133,
      "reward": 1.0034300796687603,
      "reward_std": 0.29149999690707773,
      "rewards/equation_reward_func": 0.13289434765465558,
      "rewards/format_reward_func": 0.870535722002387,
      "step": 290
    },
    {
      "completion_length": 205.84524059295654,
      "epoch": 7.1244167962674965,
      "grad_norm": 7.47841739654541,
      "kl": 92.62634913623333,
      "learning_rate": 1.4525327722875568e-07,
      "loss": 0.0926,
      "reward": 1.006636930629611,
      "reward_std": 0.29602639051154256,
      "rewards/equation_reward_func": 0.1413095265161246,
      "rewards/format_reward_func": 0.8653273889794946,
      "step": 292
    },
    {
      "completion_length": 216.39211654663086,
      "epoch": 7.174183514774494,
      "grad_norm": 60.437217712402344,
      "kl": 154.1758649945259,
      "learning_rate": 1.4199300333930515e-07,
      "loss": 0.1542,
      "reward": 0.981852700933814,
      "reward_std": 0.3057260474888608,
      "rewards/equation_reward_func": 0.1269419682212174,
      "rewards/format_reward_func": 0.8549107229337096,
      "step": 294
    },
    {
      "completion_length": 220.96652054786682,
      "epoch": 7.223950233281493,
      "grad_norm": 8.677336692810059,
      "kl": 113.40557888150215,
      "learning_rate": 1.3875515954583523e-07,
      "loss": 0.1134,
      "reward": 0.9865774074569345,
      "reward_std": 0.32082074461504817,
      "rewards/equation_reward_func": 0.13761905138380826,
      "rewards/format_reward_func": 0.8489583423361182,
      "step": 296
    },
    {
      "completion_length": 214.886164188385,
      "epoch": 7.273716951788492,
      "grad_norm": 7.040713787078857,
      "kl": 104.41407984495163,
      "learning_rate": 1.3554041825985e-07,
      "loss": 0.1044,
      "reward": 0.9882515128701925,
      "reward_std": 0.3001827346161008,
      "rewards/equation_reward_func": 0.12515625066589564,
      "rewards/format_reward_func": 0.8630952462553978,
      "step": 298
    },
    {
      "completion_length": 213.91815853118896,
      "epoch": 7.32348367029549,
      "grad_norm": 5.163806438446045,
      "kl": 110.24214600026608,
      "learning_rate": 1.323494470950949e-07,
      "loss": 0.1102,
      "reward": 0.9937500273808837,
      "reward_std": 0.29727713589090854,
      "rewards/equation_reward_func": 0.14032738481182605,
      "rewards/format_reward_func": 0.8534226277843118,
      "step": 300
    },
    {
      "completion_length": 221.51190876960754,
      "epoch": 7.3732503888024885,
      "grad_norm": 5.808466911315918,
      "kl": 105.10413581132889,
      "learning_rate": 1.2918290872891236e-07,
      "loss": 0.1051,
      "reward": 0.9996800795197487,
      "reward_std": 0.282098678057082,
      "rewards/equation_reward_func": 0.1358407783554867,
      "rewards/format_reward_func": 0.8638392938300967,
      "step": 302
    },
    {
      "completion_length": 216.43155217170715,
      "epoch": 7.423017107309486,
      "grad_norm": 5.777767658233643,
      "kl": 105.67957159876823,
      "learning_rate": 1.260414607646213e-07,
      "loss": 0.1057,
      "reward": 0.9903050791472197,
      "reward_std": 0.29644761607050896,
      "rewards/equation_reward_func": 0.12869791907723993,
      "rewards/format_reward_func": 0.861607151106,
      "step": 304
    },
    {
      "completion_length": 219.77455735206604,
      "epoch": 7.472783825816485,
      "grad_norm": 5.2907562255859375,
      "kl": 106.08304291963577,
      "learning_rate": 1.2292575559495143e-07,
      "loss": 0.1061,
      "reward": 1.0122916884720325,
      "reward_std": 0.28524951916188,
      "rewards/equation_reward_func": 0.13803571776952595,
      "rewards/format_reward_func": 0.8742559598758817,
      "step": 306
    },
    {
      "completion_length": 217.84301114082336,
      "epoch": 7.522550544323484,
      "grad_norm": 24.005651473999023,
      "kl": 124.43787896633148,
      "learning_rate": 1.1983644026655835e-07,
      "loss": 0.1244,
      "reward": 0.9638169929385185,
      "reward_std": 0.3310524779371917,
      "rewards/equation_reward_func": 0.12973958678776398,
      "rewards/format_reward_func": 0.8340773908421397,
      "step": 308
    },
    {
      "completion_length": 224.14881300926208,
      "epoch": 7.572317262830482,
      "grad_norm": 4.903006076812744,
      "kl": 116.65262049436569,
      "learning_rate": 1.1677415634565066e-07,
      "loss": 0.1167,
      "reward": 0.9956845454871655,
      "reward_std": 0.2962368195876479,
      "rewards/equation_reward_func": 0.13556547928601503,
      "rewards/format_reward_func": 0.8601190559566021,
      "step": 310
    },
    {
      "completion_length": 228.49107551574707,
      "epoch": 7.6220839813374806,
      "grad_norm": 7.504365921020508,
      "kl": 93.5646702349186,
      "learning_rate": 1.1373953978475353e-07,
      "loss": 0.0936,
      "reward": 0.9905357416719198,
      "reward_std": 0.29693792841862887,
      "rewards/equation_reward_func": 0.13116071629337966,
      "rewards/format_reward_func": 0.8593750083819032,
      "step": 312
    },
    {
      "completion_length": 210.6443498134613,
      "epoch": 7.671850699844479,
      "grad_norm": 6.315691947937012,
      "kl": 88.66525352746248,
      "learning_rate": 1.1073322079063913e-07,
      "loss": 0.0887,
      "reward": 1.0083631156012416,
      "reward_std": 0.2873585835332051,
      "rewards/equation_reward_func": 0.13782738358713686,
      "rewards/format_reward_func": 0.870535722002387,
      "step": 314
    },
    {
      "completion_length": 216.66592574119568,
      "epoch": 7.721617418351477,
      "grad_norm": 5.726399898529053,
      "kl": 94.92659878730774,
      "learning_rate": 1.0775582369344946e-07,
      "loss": 0.0949,
      "reward": 0.9947470482438803,
      "reward_std": 0.3048968045040965,
      "rewards/equation_reward_func": 0.14206845592707396,
      "rewards/format_reward_func": 0.8526785802096128,
      "step": 316
    },
    {
      "completion_length": 221.99702906608582,
      "epoch": 7.771384136858476,
      "grad_norm": 8.633785247802734,
      "kl": 127.03314569592476,
      "learning_rate": 1.0480796681704077e-07,
      "loss": 0.127,
      "reward": 0.9732217434793711,
      "reward_std": 0.3270513039897196,
      "rewards/equation_reward_func": 0.1331919669173658,
      "rewards/format_reward_func": 0.8400297714397311,
      "step": 318
    },
    {
      "completion_length": 217.00595712661743,
      "epoch": 7.821150855365475,
      "grad_norm": 10.5038480758667,
      "kl": 114.0509955137968,
      "learning_rate": 1.018902623505741e-07,
      "loss": 0.1141,
      "reward": 0.9919643187895417,
      "reward_std": 0.2979181023547426,
      "rewards/equation_reward_func": 0.1348214315949008,
      "rewards/format_reward_func": 0.8571428656578064,
      "step": 320
    },
    {
      "completion_length": 221.0297656059265,
      "epoch": 7.870917573872473,
      "grad_norm": 5.655003547668457,
      "kl": 103.35251066088676,
      "learning_rate": 9.900331622138063e-08,
      "loss": 0.1034,
      "reward": 1.003906267695129,
      "reward_std": 0.28754603571724147,
      "rewards/equation_reward_func": 0.1363467294140719,
      "rewards/format_reward_func": 0.8675595317035913,
      "step": 322
    },
    {
      "completion_length": 219.47693920135498,
      "epoch": 7.920684292379471,
      "grad_norm": 10.707879066467285,
      "kl": 114.98984357714653,
      "learning_rate": 9.614772796912681e-08,
      "loss": 0.115,
      "reward": 0.9944940665736794,
      "reward_std": 0.31038948777131736,
      "rewards/equation_reward_func": 0.1380952401086688,
      "rewards/format_reward_func": 0.8563988180831075,
      "step": 324
    },
    {
      "completion_length": 205.16146230697632,
      "epoch": 7.970451010886469,
      "grad_norm": 13.826483726501465,
      "kl": 117.3221942782402,
      "learning_rate": 9.332409062130686e-08,
      "loss": 0.1173,
      "reward": 0.9965178854763508,
      "reward_std": 0.2970969423186034,
      "rewards/equation_reward_func": 0.13788690755609423,
      "rewards/format_reward_func": 0.8586309608072042,
      "step": 326
    },
    {
      "completion_length": 221.76065625642477,
      "epoch": 8.0,
      "grad_norm": 2.785691976547241,
      "kl": 124.96122721621865,
      "learning_rate": 9.053299057008699e-08,
      "loss": 0.0742,
      "reward": 0.9635088020249417,
      "reward_std": 0.33088268221993195,
      "rewards/equation_reward_func": 0.12766917579268156,
      "rewards/format_reward_func": 0.8358396087822161,
      "step": 328
    },
    {
      "completion_length": 224.9680097103119,
      "epoch": 8.049766718506998,
      "grad_norm": 6.289979934692383,
      "kl": 101.34825879335403,
      "learning_rate": 8.777500745052743e-08,
      "loss": 0.1013,
      "reward": 1.0041592558845878,
      "reward_std": 0.2731919828802347,
      "rewards/equation_reward_func": 0.12543899181764573,
      "rewards/format_reward_func": 0.8787202453240752,
      "step": 330
    },
    {
      "completion_length": 211.30134320259094,
      "epoch": 8.099533437013998,
      "grad_norm": 8.478480339050293,
      "kl": 95.52975815534592,
      "learning_rate": 8.505071402020892e-08,
      "loss": 0.0955,
      "reward": 1.0018899161368608,
      "reward_std": 0.2962851036572829,
      "rewards/equation_reward_func": 0.13805059832520783,
      "rewards/format_reward_func": 0.8638392938300967,
      "step": 332
    },
    {
      "completion_length": 210.46205806732178,
      "epoch": 8.149300155520995,
      "grad_norm": 7.800393581390381,
      "kl": 93.41920287907124,
      "learning_rate": 8.236067604028562e-08,
      "loss": 0.0934,
      "reward": 1.0006845509633422,
      "reward_std": 0.28785590920597315,
      "rewards/equation_reward_func": 0.1361011965200305,
      "rewards/format_reward_func": 0.8645833414047956,
      "step": 334
    },
    {
      "completion_length": 213.39583683013916,
      "epoch": 8.199066874027993,
      "grad_norm": 14.319684028625488,
      "kl": 103.72171029448509,
      "learning_rate": 7.970545215799327e-08,
      "loss": 0.1037,
      "reward": 0.9992336509749293,
      "reward_std": 0.2858052357332781,
      "rewards/equation_reward_func": 0.1353943480644375,
      "rewards/format_reward_func": 0.8638392938300967,
      "step": 336
    },
    {
      "completion_length": 222.47917068004608,
      "epoch": 8.248833592534993,
      "grad_norm": 9.096053123474121,
      "kl": 112.85475918650627,
      "learning_rate": 7.708559379063204e-08,
      "loss": 0.1129,
      "reward": 0.9990848414599895,
      "reward_std": 0.28263663535472006,
      "rewards/equation_reward_func": 0.13003720622509718,
      "rewards/format_reward_func": 0.8690476268529892,
      "step": 338
    },
    {
      "completion_length": 218.63988590240479,
      "epoch": 8.29860031104199,
      "grad_norm": 6.8109869956970215,
      "kl": 129.12899693846703,
      "learning_rate": 7.45016450110534e-08,
      "loss": 0.1291,
      "reward": 0.9899181788787246,
      "reward_std": 0.29375347727909684,
      "rewards/equation_reward_func": 0.13128720642998815,
      "rewards/format_reward_func": 0.8586309608072042,
      "step": 340
    },
    {
      "completion_length": 218.00372409820557,
      "epoch": 8.348367029548989,
      "grad_norm": 14.347524642944336,
      "kl": 117.67749184370041,
      "learning_rate": 7.195414243467029e-08,
      "loss": 0.1177,
      "reward": 1.00003722589463,
      "reward_std": 0.2877940039616078,
      "rewards/equation_reward_func": 0.13619791890960187,
      "rewards/format_reward_func": 0.8638392938300967,
      "step": 342
    },
    {
      "completion_length": 216.2247062921524,
      "epoch": 8.398133748055988,
      "grad_norm": 13.07165241241455,
      "kl": 106.50291386246681,
      "learning_rate": 6.944361510801763e-08,
      "loss": 0.1065,
      "reward": 1.0014137290418148,
      "reward_std": 0.2730475334683433,
      "rewards/equation_reward_func": 0.13013393071014434,
      "rewards/format_reward_func": 0.871279769577086,
      "step": 344
    },
    {
      "completion_length": 220.52009344100952,
      "epoch": 8.447900466562986,
      "grad_norm": 6.053815841674805,
      "kl": 101.54940500855446,
      "learning_rate": 6.697058439888283e-08,
      "loss": 0.1015,
      "reward": 1.0040476340800524,
      "reward_std": 0.29517007851973176,
      "rewards/equation_reward_func": 0.14020833862014115,
      "rewards/format_reward_func": 0.8638392938300967,
      "step": 346
    },
    {
      "completion_length": 222.90997338294983,
      "epoch": 8.497667185069984,
      "grad_norm": 11.188461303710938,
      "kl": 96.22405216097832,
      "learning_rate": 6.453556388803288e-08,
      "loss": 0.0962,
      "reward": 0.9793526967987418,
      "reward_std": 0.30874833301641047,
      "rewards/equation_reward_func": 0.131882444024086,
      "rewards/format_reward_func": 0.8474702471867204,
      "step": 348
    },
    {
      "completion_length": 218.2782781124115,
      "epoch": 8.547433903576984,
      "grad_norm": 5.3335490226745605,
      "kl": 100.43345513939857,
      "learning_rate": 6.213905926255697e-08,
      "loss": 0.1004,
      "reward": 1.0058556757867336,
      "reward_std": 0.30363914044573903,
      "rewards/equation_reward_func": 0.1449925625929609,
      "rewards/format_reward_func": 0.860863103531301,
      "step": 350
    },
    {
      "completion_length": 214.927832365036,
      "epoch": 8.597200622083982,
      "grad_norm": 5.736093521118164,
      "kl": 106.57747828960419,
      "learning_rate": 5.978156821084987e-08,
      "loss": 0.1066,
      "reward": 0.9937872216105461,
      "reward_std": 0.2965135461417958,
      "rewards/equation_reward_func": 0.1373883937485516,
      "rewards/format_reward_func": 0.8563988180831075,
      "step": 352
    },
    {
      "completion_length": 209.39360523223877,
      "epoch": 8.64696734059098,
      "grad_norm": 10.456531524658203,
      "kl": 115.35611689090729,
      "learning_rate": 5.7463580319254853e-08,
      "loss": 0.1154,
      "reward": 0.9899628274142742,
      "reward_std": 0.2897207149071619,
      "rewards/equation_reward_func": 0.1305878018029034,
      "rewards/format_reward_func": 0.8593750083819032,
      "step": 354
    },
    {
      "completion_length": 215.7492594718933,
      "epoch": 8.696734059097977,
      "grad_norm": 8.979085922241211,
      "kl": 144.21495592594147,
      "learning_rate": 5.518557697039081e-08,
      "loss": 0.1442,
      "reward": 0.9727529920637608,
      "reward_std": 0.30324876098893583,
      "rewards/equation_reward_func": 0.13495535892434418,
      "rewards/format_reward_func": 0.8377976287156343,
      "step": 356
    },
    {
      "completion_length": 215.43675899505615,
      "epoch": 8.746500777604977,
      "grad_norm": 9.370060920715332,
      "kl": 118.48588564991951,
      "learning_rate": 5.294803124318145e-08,
      "loss": 0.1185,
      "reward": 1.003474730066955,
      "reward_std": 0.29565710644237697,
      "rewards/equation_reward_func": 0.14037946588359773,
      "rewards/format_reward_func": 0.8630952462553978,
      "step": 358
    },
    {
      "completion_length": 216.35863542556763,
      "epoch": 8.796267496111975,
      "grad_norm": 10.175962448120117,
      "kl": 125.35718154907227,
      "learning_rate": 5.07514078146106e-08,
      "loss": 0.1254,
      "reward": 0.9844866311177611,
      "reward_std": 0.3086923783412203,
      "rewards/equation_reward_func": 0.13106399052776396,
      "rewards/format_reward_func": 0.8534226277843118,
      "step": 360
    },
    {
      "completion_length": 216.0572965145111,
      "epoch": 8.846034214618973,
      "grad_norm": 14.969239234924316,
      "kl": 114.87217891216278,
      "learning_rate": 4.859616286322094e-08,
      "loss": 0.1149,
      "reward": 0.9952976554632187,
      "reward_std": 0.30335904110688716,
      "rewards/equation_reward_func": 0.1433630979154259,
      "rewards/format_reward_func": 0.8519345326349139,
      "step": 362
    },
    {
      "completion_length": 221.4404798746109,
      "epoch": 8.895800933125972,
      "grad_norm": 6.283931255340576,
      "kl": 110.71545603871346,
      "learning_rate": 4.648274397437829e-08,
      "loss": 0.1107,
      "reward": 0.9908779989928007,
      "reward_std": 0.3113563316874206,
      "rewards/equation_reward_func": 0.13968750124331564,
      "rewards/format_reward_func": 0.851190485060215,
      "step": 364
    },
    {
      "completion_length": 218.71875619888306,
      "epoch": 8.94556765163297,
      "grad_norm": 11.783354759216309,
      "kl": 100.00087642669678,
      "learning_rate": 4.4411590047320617e-08,
      "loss": 0.1,
      "reward": 0.982001505792141,
      "reward_std": 0.3113343908917159,
      "rewards/equation_reward_func": 0.13527529931161553,
      "rewards/format_reward_func": 0.8467261996120214,
      "step": 366
    },
    {
      "completion_length": 217.48289155960083,
      "epoch": 8.995334370139968,
      "grad_norm": 5.263770580291748,
      "kl": 103.37672221660614,
      "learning_rate": 4.2383131204010494e-08,
      "loss": 0.1034,
      "reward": 0.9819196779280901,
      "reward_std": 0.3062695120461285,
      "rewards/equation_reward_func": 0.13370536046568304,
      "rewards/format_reward_func": 0.8482142947614193,
      "step": 368
    },
    {
      "completion_length": 214.46491683156867,
      "epoch": 9.024883359253499,
      "grad_norm": 12.257465362548828,
      "kl": 99.55310023458381,
      "learning_rate": 4.039778869981064e-08,
      "loss": 0.0591,
      "reward": 0.9891980086502276,
      "reward_std": 0.29844278195186663,
      "rewards/equation_reward_func": 0.13456140644848347,
      "rewards/format_reward_func": 0.8546366001430311,
      "step": 370
    },
    {
      "completion_length": 204.44047844409943,
      "epoch": 9.074650077760497,
      "grad_norm": 6.687661647796631,
      "kl": 102.91622030735016,
      "learning_rate": 3.845597483600049e-08,
      "loss": 0.1029,
      "reward": 1.00175597704947,
      "reward_std": 0.28237274824641645,
      "rewards/equation_reward_func": 0.1341964314924553,
      "rewards/format_reward_func": 0.8675595317035913,
      "step": 372
    },
    {
      "completion_length": 215.5305094718933,
      "epoch": 9.124416796267496,
      "grad_norm": 6.855081558227539,
      "kl": 114.0059272646904,
      "learning_rate": 3.655809287415284e-08,
      "loss": 0.114,
      "reward": 0.9824330499395728,
      "reward_std": 0.31068597990088165,
      "rewards/equation_reward_func": 0.1319866105914116,
      "rewards/format_reward_func": 0.8504464374855161,
      "step": 374
    },
    {
      "completion_length": 213.50446844100952,
      "epoch": 9.174183514774494,
      "grad_norm": 8.869424819946289,
      "kl": 116.45409867167473,
      "learning_rate": 3.4704536952387285e-08,
      "loss": 0.1165,
      "reward": 0.996569961309433,
      "reward_std": 0.29541023285128176,
      "rewards/equation_reward_func": 0.13719494419638067,
      "rewards/format_reward_func": 0.8593750083819032,
      "step": 376
    },
    {
      "completion_length": 209.6510455608368,
      "epoch": 9.223950233281492,
      "grad_norm": 5.935553550720215,
      "kl": 105.62733614444733,
      "learning_rate": 3.2895692003518575e-08,
      "loss": 0.1056,
      "reward": 1.018154788762331,
      "reward_std": 0.2831184302922338,
      "rewards/equation_reward_func": 0.1409226229880005,
      "rewards/format_reward_func": 0.8772321501746774,
      "step": 378
    },
    {
      "completion_length": 211.81324791908264,
      "epoch": 9.273716951788492,
      "grad_norm": 7.783820629119873,
      "kl": 121.14890000224113,
      "learning_rate": 3.113193367511635e-08,
      "loss": 0.1211,
      "reward": 0.9982961546629667,
      "reward_std": 0.2949765467783436,
      "rewards/equation_reward_func": 0.13296875241212547,
      "rewards/format_reward_func": 0.8653273889794946,
      "step": 380
    },
    {
      "completion_length": 216.39137244224548,
      "epoch": 9.32348367029549,
      "grad_norm": 6.670078754425049,
      "kl": 126.6070476770401,
      "learning_rate": 2.9413628251493934e-08,
      "loss": 0.1266,
      "reward": 1.0011904956772923,
      "reward_std": 0.2955527463927865,
      "rewards/equation_reward_func": 0.13288690755143762,
      "rewards/format_reward_func": 0.8683035792782903,
      "step": 382
    },
    {
      "completion_length": 212.9427134990692,
      "epoch": 9.373250388802488,
      "grad_norm": 5.376703262329102,
      "kl": 127.98397767543793,
      "learning_rate": 2.774113257764066e-08,
      "loss": 0.128,
      "reward": 0.982187514193356,
      "reward_std": 0.32169102667830884,
      "rewards/equation_reward_func": 0.13620536087546498,
      "rewards/format_reward_func": 0.8459821520373225,
      "step": 384
    },
    {
      "completion_length": 211.45015382766724,
      "epoch": 9.423017107309487,
      "grad_norm": 8.18055248260498,
      "kl": 119.37251782417297,
      "learning_rate": 2.611479398511518e-08,
      "loss": 0.1194,
      "reward": 0.9903794955462217,
      "reward_std": 0.2968536267289892,
      "rewards/equation_reward_func": 0.13026041921693832,
      "rewards/format_reward_func": 0.8601190559566021,
      "step": 386
    },
    {
      "completion_length": 204.16592621803284,
      "epoch": 9.472783825816485,
      "grad_norm": 7.270776748657227,
      "kl": 101.86209693551064,
      "learning_rate": 2.4534950219914057e-08,
      "loss": 0.1019,
      "reward": 0.9984970455989242,
      "reward_std": 0.29997857997659594,
      "rewards/equation_reward_func": 0.13837797672022134,
      "rewards/format_reward_func": 0.8601190559566021,
      "step": 388
    },
    {
      "completion_length": 214.3392894268036,
      "epoch": 9.522550544323483,
      "grad_norm": 7.736979007720947,
      "kl": 105.20895183086395,
      "learning_rate": 2.300192937233128e-08,
      "loss": 0.1052,
      "reward": 0.9999851500615478,
      "reward_std": 0.28676424000877887,
      "rewards/equation_reward_func": 0.1346577397780493,
      "rewards/format_reward_func": 0.8653273889794946,
      "step": 390
    },
    {
      "completion_length": 210.4516406059265,
      "epoch": 9.572317262830483,
      "grad_norm": 6.528294563293457,
      "kl": 106.73470288515091,
      "learning_rate": 2.1516049808822935e-08,
      "loss": 0.1067,
      "reward": 0.9907887103036046,
      "reward_std": 0.30266366386786103,
      "rewards/equation_reward_func": 0.13438988395500928,
      "rewards/format_reward_func": 0.8563988180831075,
      "step": 392
    },
    {
      "completion_length": 217.03423070907593,
      "epoch": 9.62208398133748,
      "grad_norm": 8.291276931762695,
      "kl": 112.02839949727058,
      "learning_rate": 2.007762010589098e-08,
      "loss": 0.112,
      "reward": 0.9856399018317461,
      "reward_std": 0.29987687326502055,
      "rewards/equation_reward_func": 0.13519345549866557,
      "rewards/format_reward_func": 0.8504464374855161,
      "step": 394
    },
    {
      "completion_length": 212.06771171092987,
      "epoch": 9.671850699844478,
      "grad_norm": 23.273250579833984,
      "kl": 114.7417793571949,
      "learning_rate": 1.8686938986000627e-08,
      "loss": 0.1147,
      "reward": 1.0074405102059245,
      "reward_std": 0.2830245600780472,
      "rewards/equation_reward_func": 0.1391369072953239,
      "rewards/format_reward_func": 0.8683035792782903,
      "step": 396
    },
    {
      "completion_length": 208.95982575416565,
      "epoch": 9.721617418351478,
      "grad_norm": 8.359769821166992,
      "kl": 105.31409025192261,
      "learning_rate": 1.734429525554365e-08,
      "loss": 0.1053,
      "reward": 0.9976785909384489,
      "reward_std": 0.2738517120014876,
      "rewards/equation_reward_func": 0.13309523789212108,
      "rewards/format_reward_func": 0.8645833414047956,
      "step": 398
    },
    {
      "completion_length": 220.13095664978027,
      "epoch": 9.771384136858476,
      "grad_norm": 5.326533317565918,
      "kl": 104.11170220375061,
      "learning_rate": 1.604996774486145e-08,
      "loss": 0.1041,
      "reward": 0.9921280033886433,
      "reward_std": 0.2880916509311646,
      "rewards/equation_reward_func": 0.13796131452545524,
      "rewards/format_reward_func": 0.8541666753590107,
      "step": 400
    },
    {
      "completion_length": 203.88393235206604,
      "epoch": 9.821150855365474,
      "grad_norm": 5.928884029388428,
      "kl": 100.95771364122629,
      "learning_rate": 1.4804225250339281e-08,
      "loss": 0.101,
      "reward": 0.988816992379725,
      "reward_std": 0.31290939077734947,
      "rewards/equation_reward_func": 0.13613839482422918,
      "rewards/format_reward_func": 0.8526785802096128,
      "step": 402
    },
    {
      "completion_length": 217.1480677127838,
      "epoch": 9.870917573872473,
      "grad_norm": 4.631835460662842,
      "kl": 93.6852602660656,
      "learning_rate": 1.360732647858498e-08,
      "loss": 0.0937,
      "reward": 0.9889360275119543,
      "reward_std": 0.286705979029648,
      "rewards/equation_reward_func": 0.13030506228096783,
      "rewards/format_reward_func": 0.8586309608072042,
      "step": 404
    },
    {
      "completion_length": 222.09077751636505,
      "epoch": 9.920684292379471,
      "grad_norm": 5.06677770614624,
      "kl": 95.56222039461136,
      "learning_rate": 1.2459519992702311e-08,
      "loss": 0.0956,
      "reward": 1.0050446651875973,
      "reward_std": 0.2898379734251648,
      "rewards/equation_reward_func": 0.1322767889359966,
      "rewards/format_reward_func": 0.8727678647264838,
      "step": 406
    },
    {
      "completion_length": 223.86830830574036,
      "epoch": 9.97045101088647,
      "grad_norm": 4.548260688781738,
      "kl": 99.37585133314133,
      "learning_rate": 1.1361044160671629e-08,
      "loss": 0.0994,
      "reward": 1.0121131222695112,
      "reward_std": 0.29200624965596944,
      "rewards/equation_reward_func": 0.13934524141950533,
      "rewards/format_reward_func": 0.8727678647264838,
      "step": 408
    },
    {
      "completion_length": 207.73434006540398,
      "epoch": 10.0,
      "grad_norm": 1.7612920999526978,
      "kl": 90.04842346592953,
      "learning_rate": 1.0312127105846947e-08,
      "loss": 0.0535,
      "reward": 1.0120050326773995,
      "reward_std": 0.2693111804362975,
      "rewards/equation_reward_func": 0.146090227718416,
      "rewards/format_reward_func": 0.8659147949595201,
      "step": 410
    },
    {
      "completion_length": 212.9203917980194,
      "epoch": 10.049766718506998,
      "grad_norm": 9.370685577392578,
      "kl": 89.39757791161537,
      "learning_rate": 9.312986659581301e-09,
      "loss": 0.0894,
      "reward": 1.0107515146955848,
      "reward_std": 0.27970271767117083,
      "rewards/equation_reward_func": 0.13351934927050024,
      "rewards/format_reward_func": 0.8772321501746774,
      "step": 412
    },
    {
      "completion_length": 216.9858672618866,
      "epoch": 10.099533437013998,
      "grad_norm": 5.712127208709717,
      "kl": 97.55916061997414,
      "learning_rate": 8.363830315988945e-09,
      "loss": 0.0976,
      "reward": 1.0012500202283263,
      "reward_std": 0.2857985398732126,
      "rewards/equation_reward_func": 0.13592262053862214,
      "rewards/format_reward_func": 0.8653273889794946,
      "step": 414
    },
    {
      "completion_length": 206.04910945892334,
      "epoch": 10.149300155520995,
      "grad_norm": 8.428742408752441,
      "kl": 103.54971778392792,
      "learning_rate": 7.46485518885462e-09,
      "loss": 0.1035,
      "reward": 0.9935044907033443,
      "reward_std": 0.29330903699155897,
      "rewards/equation_reward_func": 0.13933780090883374,
      "rewards/format_reward_func": 0.8541666753590107,
      "step": 416
    },
    {
      "completion_length": 216.40104627609253,
      "epoch": 10.199066874027993,
      "grad_norm": 7.276556968688965,
      "kl": 96.23605534434319,
      "learning_rate": 6.616247970698319e-09,
      "loss": 0.0962,
      "reward": 1.003549126908183,
      "reward_std": 0.28676802827976644,
      "rewards/equation_reward_func": 0.13822172884829342,
      "rewards/format_reward_func": 0.8653273889794946,
      "step": 418
    },
    {
      "completion_length": 217.37574791908264,
      "epoch": 10.248833592534993,
      "grad_norm": 7.763327598571777,
      "kl": 105.16229781508446,
      "learning_rate": 5.8181848940044855e-09,
      "loss": 0.1052,
      "reward": 1.0146056832745671,
      "reward_std": 0.2754618857288733,
      "rewards/equation_reward_func": 0.13886161183472723,
      "rewards/format_reward_func": 0.8757440550252795,
      "step": 420
    },
    {
      "completion_length": 205.62202739715576,
      "epoch": 10.29860031104199,
      "grad_norm": 10.51353931427002,
      "kl": 95.59216642379761,
      "learning_rate": 5.070831694623135e-09,
      "loss": 0.0956,
      "reward": 0.999791688285768,
      "reward_std": 0.3051185356453061,
      "rewards/equation_reward_func": 0.13818452623672783,
      "rewards/format_reward_func": 0.861607151106,
      "step": 422
    },
    {
      "completion_length": 217.62425827980042,
      "epoch": 10.348367029548989,
      "grad_norm": 7.500198841094971,
      "kl": 105.92476946115494,
      "learning_rate": 4.374343577351336e-09,
      "loss": 0.1059,
      "reward": 0.9995759120211005,
      "reward_std": 0.28294855484273285,
      "rewards/equation_reward_func": 0.13648065709276125,
      "rewards/format_reward_func": 0.8630952462553978,
      "step": 424
    },
    {
      "completion_length": 218.08780217170715,
      "epoch": 10.398133748055988,
      "grad_norm": 5.859328269958496,
      "kl": 112.09179133176804,
      "learning_rate": 3.7288651837012745e-09,
      "loss": 0.1121,
      "reward": 0.9871056908741593,
      "reward_std": 0.305632587405853,
      "rewards/equation_reward_func": 0.13442708633374423,
      "rewards/format_reward_func": 0.8526785802096128,
      "step": 426
    },
    {
      "completion_length": 207.83631467819214,
      "epoch": 10.447900466562986,
      "grad_norm": 8.897568702697754,
      "kl": 107.37559473514557,
      "learning_rate": 3.134530561862081e-09,
      "loss": 0.1074,
      "reward": 0.9991220543161035,
      "reward_std": 0.2914143274538219,
      "rewards/equation_reward_func": 0.1375148834194988,
      "rewards/format_reward_func": 0.861607151106,
      "step": 428
    },
    {
      "completion_length": 212.729914188385,
      "epoch": 10.497667185069984,
      "grad_norm": 8.106195449829102,
      "kl": 115.35286539793015,
      "learning_rate": 2.5914631388619103e-09,
      "loss": 0.1154,
      "reward": 0.9826562711969018,
      "reward_std": 0.3164314874447882,
      "rewards/equation_reward_func": 0.13146577589213848,
      "rewards/format_reward_func": 0.851190485060215,
      "step": 430
    },
    {
      "completion_length": 215.90774369239807,
      "epoch": 10.547433903576984,
      "grad_norm": 13.18673038482666,
      "kl": 112.25744771957397,
      "learning_rate": 2.0997756949353297e-09,
      "loss": 0.1123,
      "reward": 1.0021726433187723,
      "reward_std": 0.29007076751440763,
      "rewards/equation_reward_func": 0.1346130990423262,
      "rewards/format_reward_func": 0.8675595317035913,
      "step": 432
    },
    {
      "completion_length": 209.4575924873352,
      "epoch": 10.597200622083982,
      "grad_norm": 7.412752628326416,
      "kl": 92.43682231009007,
      "learning_rate": 1.6595703401020844e-09,
      "loss": 0.0924,
      "reward": 1.0252753188833594,
      "reward_std": 0.26438383577624336,
      "rewards/equation_reward_func": 0.1361383959883824,
      "rewards/format_reward_func": 0.8891369113698602,
      "step": 434
    },
    {
      "completion_length": 212.07961678504944,
      "epoch": 10.64696734059098,
      "grad_norm": 9.115935325622559,
      "kl": 98.72861561179161,
      "learning_rate": 1.2709384929615596e-09,
      "loss": 0.0987,
      "reward": 1.0122768012806773,
      "reward_std": 0.2858473288360983,
      "rewards/equation_reward_func": 0.14174107427243143,
      "rewards/format_reward_func": 0.870535722002387,
      "step": 436
    },
    {
      "completion_length": 213.41666972637177,
      "epoch": 10.696734059097977,
      "grad_norm": 7.424521446228027,
      "kl": 101.00204753875732,
      "learning_rate": 9.339608617077165e-10,
      "loss": 0.101,
      "reward": 1.0160119328647852,
      "reward_std": 0.27073649514932185,
      "rewards/equation_reward_func": 0.13654762151418254,
      "rewards/format_reward_func": 0.8794642928987741,
      "step": 438
    },
    {
      "completion_length": 214.2976235151291,
      "epoch": 10.746500777604977,
      "grad_norm": 6.853705883026123,
      "kl": 103.29757300019264,
      "learning_rate": 6.487074273681114e-10,
      "loss": 0.1033,
      "reward": 0.9976488519459963,
      "reward_std": 0.2918668915517628,
      "rewards/equation_reward_func": 0.13604166894219816,
      "rewards/format_reward_func": 0.861607151106,
      "step": 440
    },
    {
      "completion_length": 212.39434838294983,
      "epoch": 10.796267496111975,
      "grad_norm": 5.543258190155029,
      "kl": 106.93525078892708,
      "learning_rate": 4.152374292708538e-10,
      "loss": 0.1069,
      "reward": 1.0103422962129116,
      "reward_std": 0.28469230665359646,
      "rewards/equation_reward_func": 0.13683036074507982,
      "rewards/format_reward_func": 0.8735119123011827,
      "step": 442
    },
    {
      "completion_length": 213.31473565101624,
      "epoch": 10.846034214618973,
      "grad_norm": 7.63397741317749,
      "kl": 108.80926709622145,
      "learning_rate": 2.3359935274214204e-10,
      "loss": 0.1088,
      "reward": 0.9913095412775874,
      "reward_std": 0.2936742458259687,
      "rewards/equation_reward_func": 0.13044643041212112,
      "rewards/format_reward_func": 0.860863103531301,
      "step": 444
    },
    {
      "completion_length": 217.44345545768738,
      "epoch": 10.895800933125972,
      "grad_norm": 6.197990894317627,
      "kl": 106.08692070841789,
      "learning_rate": 1.0383091903720665e-10,
      "loss": 0.1061,
      "reward": 0.994486641138792,
      "reward_std": 0.29480275977402925,
      "rewards/equation_reward_func": 0.1373437523143366,
      "rewards/format_reward_func": 0.8571428656578064,
      "step": 446
    },
    {
      "completion_length": 215.1421172618866,
      "epoch": 10.94556765163297,
      "grad_norm": 13.626302719116211,
      "kl": 108.21286904811859,
      "learning_rate": 2.595907750671533e-11,
      "loss": 0.1082,
      "reward": 1.0048437686637044,
      "reward_std": 0.28983362135477364,
      "rewards/equation_reward_func": 0.1410044669173658,
      "rewards/format_reward_func": 0.8638392938300967,
      "step": 448
    },
    {
      "completion_length": 212.74553871154785,
      "epoch": 10.995334370139968,
      "grad_norm": 12.845934867858887,
      "kl": 108.94018495082855,
      "learning_rate": 0.0,
      "loss": 0.1089,
      "reward": 1.0015253257006407,
      "reward_std": 0.2928236851003021,
      "rewards/equation_reward_func": 0.13619791949167848,
      "rewards/format_reward_func": 0.8653273889794946,
      "step": 450
    },
    {
      "epoch": 10.995334370139968,
      "step": 450,
      "total_flos": 0.0,
      "train_loss": 0.10479868455554182,
      "train_runtime": 84441.5126,
      "train_samples_per_second": 1.194,
      "train_steps_per_second": 0.005
    }
  ],
  "logging_steps": 2,
  "max_steps": 450,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "save_steps": 25,
  "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": 1,
  "trial_name": null,
  "trial_params": null
}