{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7640878701050621,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0038204393505253103,
      "grad_norm": 18.127355575561523,
      "learning_rate": 4.917142857142858e-06,
      "loss": 1.314,
      "mean_token_accuracy": 0.8176221281290055,
      "num_tokens": 186088.0,
      "step": 5
    },
    {
      "epoch": 0.007640878701050621,
      "grad_norm": 6.039257049560547,
      "learning_rate": 1.106357142857143e-05,
      "loss": 0.5882,
      "mean_token_accuracy": 0.868802610039711,
      "num_tokens": 371617.0,
      "step": 10
    },
    {
      "epoch": 0.011461318051575931,
      "grad_norm": 2.100287437438965,
      "learning_rate": 1.7210000000000003e-05,
      "loss": 0.2617,
      "mean_token_accuracy": 0.9358268558979035,
      "num_tokens": 561042.0,
      "step": 15
    },
    {
      "epoch": 0.015281757402101241,
      "grad_norm": 2.6150949001312256,
      "learning_rate": 2.3356428571428574e-05,
      "loss": 0.1398,
      "mean_token_accuracy": 0.9541143894195556,
      "num_tokens": 753311.0,
      "step": 20
    },
    {
      "epoch": 0.019102196752626553,
      "grad_norm": 1.2212921380996704,
      "learning_rate": 2.950285714285715e-05,
      "loss": 0.1481,
      "mean_token_accuracy": 0.9479759961366654,
      "num_tokens": 940411.0,
      "step": 25
    },
    {
      "epoch": 0.022922636103151862,
      "grad_norm": 0.837933361530304,
      "learning_rate": 3.5649285714285724e-05,
      "loss": 0.1486,
      "mean_token_accuracy": 0.950935247540474,
      "num_tokens": 1126815.0,
      "step": 30
    },
    {
      "epoch": 0.026743075453677174,
      "grad_norm": 0.8025307655334473,
      "learning_rate": 4.179571428571429e-05,
      "loss": 0.1369,
      "mean_token_accuracy": 0.9541471898555756,
      "num_tokens": 1313603.0,
      "step": 35
    },
    {
      "epoch": 0.030563514804202482,
      "grad_norm": 1.4238449335098267,
      "learning_rate": 4.3024809062069924e-05,
      "loss": 0.108,
      "mean_token_accuracy": 0.9628856807947159,
      "num_tokens": 1501890.0,
      "step": 40
    },
    {
      "epoch": 0.034383954154727794,
      "grad_norm": 0.9520809650421143,
      "learning_rate": 4.302403338447426e-05,
      "loss": 0.1242,
      "mean_token_accuracy": 0.9570747315883636,
      "num_tokens": 1689350.0,
      "step": 45
    },
    {
      "epoch": 0.038204393505253106,
      "grad_norm": 1.042546033859253,
      "learning_rate": 4.3022661062256383e-05,
      "loss": 0.1421,
      "mean_token_accuracy": 0.9494232147932052,
      "num_tokens": 1877717.0,
      "step": 50
    },
    {
      "epoch": 0.04202483285577841,
      "grad_norm": 0.5485150814056396,
      "learning_rate": 4.3020692146167425e-05,
      "loss": 0.1558,
      "mean_token_accuracy": 0.9491366922855378,
      "num_tokens": 2061566.0,
      "step": 55
    },
    {
      "epoch": 0.045845272206303724,
      "grad_norm": 0.9261055588722229,
      "learning_rate": 4.30181267090217e-05,
      "loss": 0.1315,
      "mean_token_accuracy": 0.9554131269454956,
      "num_tokens": 2248287.0,
      "step": 60
    },
    {
      "epoch": 0.049665711556829036,
      "grad_norm": 0.6507024168968201,
      "learning_rate": 4.301496484569403e-05,
      "loss": 0.1308,
      "mean_token_accuracy": 0.9518629878759384,
      "num_tokens": 2436238.0,
      "step": 65
    },
    {
      "epoch": 0.05348615090735435,
      "grad_norm": 0.6103971004486084,
      "learning_rate": 4.3011206673116235e-05,
      "loss": 0.1091,
      "mean_token_accuracy": 0.9624674707651139,
      "num_tokens": 2625894.0,
      "step": 70
    },
    {
      "epoch": 0.05730659025787966,
      "grad_norm": 0.5037476420402527,
      "learning_rate": 4.300685233027281e-05,
      "loss": 0.1056,
      "mean_token_accuracy": 0.964278519153595,
      "num_tokens": 2811213.0,
      "step": 75
    },
    {
      "epoch": 0.061127029608404965,
      "grad_norm": 0.5023500919342041,
      "learning_rate": 4.300190197819575e-05,
      "loss": 0.1071,
      "mean_token_accuracy": 0.960286995768547,
      "num_tokens": 2997430.0,
      "step": 80
    },
    {
      "epoch": 0.06494746895893028,
      "grad_norm": 0.6508046388626099,
      "learning_rate": 4.299635579995865e-05,
      "loss": 0.1425,
      "mean_token_accuracy": 0.9518447428941726,
      "num_tokens": 3181763.0,
      "step": 85
    },
    {
      "epoch": 0.06876790830945559,
      "grad_norm": 0.3975546061992645,
      "learning_rate": 4.299021400066991e-05,
      "loss": 0.1163,
      "mean_token_accuracy": 0.9605496257543564,
      "num_tokens": 3368774.0,
      "step": 90
    },
    {
      "epoch": 0.0725883476599809,
      "grad_norm": 0.32393020391464233,
      "learning_rate": 4.2983476807465125e-05,
      "loss": 0.1062,
      "mean_token_accuracy": 0.960227832198143,
      "num_tokens": 3554803.0,
      "step": 95
    },
    {
      "epoch": 0.07640878701050621,
      "grad_norm": 0.42871594429016113,
      "learning_rate": 4.2976144469498725e-05,
      "loss": 0.1207,
      "mean_token_accuracy": 0.9562751889228821,
      "num_tokens": 3748616.0,
      "step": 100
    },
    {
      "epoch": 0.08022922636103152,
      "grad_norm": 0.5176600217819214,
      "learning_rate": 4.296821725793474e-05,
      "loss": 0.1143,
      "mean_token_accuracy": 0.9597778886556625,
      "num_tokens": 3937866.0,
      "step": 105
    },
    {
      "epoch": 0.08404966571155682,
      "grad_norm": 0.5702073574066162,
      "learning_rate": 4.295969546593677e-05,
      "loss": 0.0858,
      "mean_token_accuracy": 0.9671787112951279,
      "num_tokens": 4127323.0,
      "step": 110
    },
    {
      "epoch": 0.08787010506208214,
      "grad_norm": 0.7516563534736633,
      "learning_rate": 4.2950579408657155e-05,
      "loss": 0.089,
      "mean_token_accuracy": 0.9676755845546723,
      "num_tokens": 4317503.0,
      "step": 115
    },
    {
      "epoch": 0.09169054441260745,
      "grad_norm": 0.4344717860221863,
      "learning_rate": 4.2940869423225314e-05,
      "loss": 0.1057,
      "mean_token_accuracy": 0.9625389873981476,
      "num_tokens": 4504363.0,
      "step": 120
    },
    {
      "epoch": 0.09551098376313276,
      "grad_norm": 0.5618029236793518,
      "learning_rate": 4.293056586873527e-05,
      "loss": 0.1267,
      "mean_token_accuracy": 0.9540828675031662,
      "num_tokens": 4691830.0,
      "step": 125
    },
    {
      "epoch": 0.09933142311365807,
      "grad_norm": 0.3641500473022461,
      "learning_rate": 4.291966912623238e-05,
      "loss": 0.1206,
      "mean_token_accuracy": 0.9560654520988464,
      "num_tokens": 4877587.0,
      "step": 130
    },
    {
      "epoch": 0.10315186246418338,
      "grad_norm": 0.34205666184425354,
      "learning_rate": 4.290817959869925e-05,
      "loss": 0.0934,
      "mean_token_accuracy": 0.9666131526231766,
      "num_tokens": 5064288.0,
      "step": 135
    },
    {
      "epoch": 0.1069723018147087,
      "grad_norm": 0.4647500813007355,
      "learning_rate": 4.289609771104081e-05,
      "loss": 0.1146,
      "mean_token_accuracy": 0.9636133700609207,
      "num_tokens": 5253227.0,
      "step": 140
    },
    {
      "epoch": 0.11079274116523401,
      "grad_norm": 0.5485135912895203,
      "learning_rate": 4.2883423910068604e-05,
      "loss": 0.1084,
      "mean_token_accuracy": 0.9591280817985535,
      "num_tokens": 5444545.0,
      "step": 145
    },
    {
      "epoch": 0.11461318051575932,
      "grad_norm": 0.45034313201904297,
      "learning_rate": 4.2870158664484296e-05,
      "loss": 0.1159,
      "mean_token_accuracy": 0.9563294798135757,
      "num_tokens": 5627403.0,
      "step": 150
    },
    {
      "epoch": 0.11843361986628462,
      "grad_norm": 0.38060814142227173,
      "learning_rate": 4.285630246486231e-05,
      "loss": 0.1257,
      "mean_token_accuracy": 0.9533220291137695,
      "num_tokens": 5811738.0,
      "step": 155
    },
    {
      "epoch": 0.12225405921680993,
      "grad_norm": 0.28782889246940613,
      "learning_rate": 4.284185582363168e-05,
      "loss": 0.1061,
      "mean_token_accuracy": 0.9595954120159149,
      "num_tokens": 5999657.0,
      "step": 160
    },
    {
      "epoch": 0.12607449856733524,
      "grad_norm": 0.39637407660484314,
      "learning_rate": 4.282681927505712e-05,
      "loss": 0.079,
      "mean_token_accuracy": 0.9714627534151077,
      "num_tokens": 6188086.0,
      "step": 165
    },
    {
      "epoch": 0.12989493791786055,
      "grad_norm": 0.4155080020427704,
      "learning_rate": 4.281119337521925e-05,
      "loss": 0.1205,
      "mean_token_accuracy": 0.9555595755577088,
      "num_tokens": 6375110.0,
      "step": 170
    },
    {
      "epoch": 0.13371537726838587,
      "grad_norm": 0.31320029497146606,
      "learning_rate": 4.279497870199406e-05,
      "loss": 0.1124,
      "mean_token_accuracy": 0.9573060631752014,
      "num_tokens": 6564338.0,
      "step": 175
    },
    {
      "epoch": 0.13753581661891118,
      "grad_norm": 0.43746769428253174,
      "learning_rate": 4.277817585503149e-05,
      "loss": 0.0897,
      "mean_token_accuracy": 0.9692918300628662,
      "num_tokens": 6755541.0,
      "step": 180
    },
    {
      "epoch": 0.1413562559694365,
      "grad_norm": 0.4989294111728668,
      "learning_rate": 4.27607854557333e-05,
      "loss": 0.0967,
      "mean_token_accuracy": 0.9664305448532104,
      "num_tokens": 6945428.0,
      "step": 185
    },
    {
      "epoch": 0.1451766953199618,
      "grad_norm": 0.3684060871601105,
      "learning_rate": 4.2742808147230056e-05,
      "loss": 0.0972,
      "mean_token_accuracy": 0.9671982109546662,
      "num_tokens": 7129650.0,
      "step": 190
    },
    {
      "epoch": 0.1489971346704871,
      "grad_norm": 0.6600679159164429,
      "learning_rate": 4.2724244594357395e-05,
      "loss": 0.1139,
      "mean_token_accuracy": 0.9573622584342957,
      "num_tokens": 7313192.0,
      "step": 195
    },
    {
      "epoch": 0.15281757402101243,
      "grad_norm": 0.8039980530738831,
      "learning_rate": 4.270509548363136e-05,
      "loss": 0.0904,
      "mean_token_accuracy": 0.9672283232212067,
      "num_tokens": 7507495.0,
      "step": 200
    },
    {
      "epoch": 0.15663801337153774,
      "grad_norm": 0.3570236265659332,
      "learning_rate": 4.268536152322307e-05,
      "loss": 0.1242,
      "mean_token_accuracy": 0.9536033749580384,
      "num_tokens": 7694148.0,
      "step": 205
    },
    {
      "epoch": 0.16045845272206305,
      "grad_norm": 0.3374585509300232,
      "learning_rate": 4.266504344293253e-05,
      "loss": 0.0934,
      "mean_token_accuracy": 0.9666879683732986,
      "num_tokens": 7885095.0,
      "step": 210
    },
    {
      "epoch": 0.16427889207258833,
      "grad_norm": 0.6489589810371399,
      "learning_rate": 4.264414199416162e-05,
      "loss": 0.0928,
      "mean_token_accuracy": 0.9649216860532761,
      "num_tokens": 8075470.0,
      "step": 215
    },
    {
      "epoch": 0.16809933142311365,
      "grad_norm": 0.41782528162002563,
      "learning_rate": 4.26226579498863e-05,
      "loss": 0.0994,
      "mean_token_accuracy": 0.9623035788536072,
      "num_tokens": 8261926.0,
      "step": 220
    },
    {
      "epoch": 0.17191977077363896,
      "grad_norm": 0.39845508337020874,
      "learning_rate": 4.260059210462805e-05,
      "loss": 0.0856,
      "mean_token_accuracy": 0.9727312207221985,
      "num_tokens": 8449880.0,
      "step": 225
    },
    {
      "epoch": 0.17574021012416427,
      "grad_norm": 0.39581161737442017,
      "learning_rate": 4.257794527442448e-05,
      "loss": 0.0949,
      "mean_token_accuracy": 0.966069158911705,
      "num_tokens": 8637769.0,
      "step": 230
    },
    {
      "epoch": 0.17956064947468958,
      "grad_norm": 0.4898621737957001,
      "learning_rate": 4.2554718296799135e-05,
      "loss": 0.0906,
      "mean_token_accuracy": 0.9662930727005005,
      "num_tokens": 8824218.0,
      "step": 235
    },
    {
      "epoch": 0.1833810888252149,
      "grad_norm": 0.500076949596405,
      "learning_rate": 4.2530912030730534e-05,
      "loss": 0.1005,
      "mean_token_accuracy": 0.9624428898096085,
      "num_tokens": 9005169.0,
      "step": 240
    },
    {
      "epoch": 0.1872015281757402,
      "grad_norm": 0.33393213152885437,
      "learning_rate": 4.2506527356620404e-05,
      "loss": 0.1135,
      "mean_token_accuracy": 0.9580814898014068,
      "num_tokens": 9194770.0,
      "step": 245
    },
    {
      "epoch": 0.19102196752626552,
      "grad_norm": 0.329088032245636,
      "learning_rate": 4.248156517626113e-05,
      "loss": 0.0926,
      "mean_token_accuracy": 0.9639642715454102,
      "num_tokens": 9387565.0,
      "step": 250
    },
    {
      "epoch": 0.19484240687679083,
      "grad_norm": 0.26629024744033813,
      "learning_rate": 4.245602641280239e-05,
      "loss": 0.0796,
      "mean_token_accuracy": 0.9702959716320038,
      "num_tokens": 9576874.0,
      "step": 255
    },
    {
      "epoch": 0.19866284622731614,
      "grad_norm": 0.3390871584415436,
      "learning_rate": 4.2429912010716984e-05,
      "loss": 0.0949,
      "mean_token_accuracy": 0.9658147364854812,
      "num_tokens": 9767575.0,
      "step": 260
    },
    {
      "epoch": 0.20248328557784145,
      "grad_norm": 0.26101455092430115,
      "learning_rate": 4.2403222935766006e-05,
      "loss": 0.0871,
      "mean_token_accuracy": 0.9684312403202057,
      "num_tokens": 9955345.0,
      "step": 265
    },
    {
      "epoch": 0.20630372492836677,
      "grad_norm": 0.24978765845298767,
      "learning_rate": 4.237596017496301e-05,
      "loss": 0.1029,
      "mean_token_accuracy": 0.9638839602470398,
      "num_tokens": 10141192.0,
      "step": 270
    },
    {
      "epoch": 0.21012416427889208,
      "grad_norm": 0.30993443727493286,
      "learning_rate": 4.2348124736537575e-05,
      "loss": 0.0932,
      "mean_token_accuracy": 0.9692367345094681,
      "num_tokens": 10327327.0,
      "step": 275
    },
    {
      "epoch": 0.2139446036294174,
      "grad_norm": 0.22442157566547394,
      "learning_rate": 4.231971764989802e-05,
      "loss": 0.0953,
      "mean_token_accuracy": 0.9625919789075852,
      "num_tokens": 10515586.0,
      "step": 280
    },
    {
      "epoch": 0.2177650429799427,
      "grad_norm": 1.1560133695602417,
      "learning_rate": 4.229073996559331e-05,
      "loss": 0.0991,
      "mean_token_accuracy": 0.9642855018377304,
      "num_tokens": 10705569.0,
      "step": 285
    },
    {
      "epoch": 0.22158548233046801,
      "grad_norm": 0.34281474351882935,
      "learning_rate": 4.226119275527421e-05,
      "loss": 0.1043,
      "mean_token_accuracy": 0.9594624757766723,
      "num_tokens": 10890731.0,
      "step": 290
    },
    {
      "epoch": 0.22540592168099333,
      "grad_norm": 0.26781439781188965,
      "learning_rate": 4.2231077111653674e-05,
      "loss": 0.0891,
      "mean_token_accuracy": 0.9683564990758896,
      "num_tokens": 11081280.0,
      "step": 295
    },
    {
      "epoch": 0.22922636103151864,
      "grad_norm": 0.25002506375312805,
      "learning_rate": 4.220039414846639e-05,
      "loss": 0.0891,
      "mean_token_accuracy": 0.9684525817632675,
      "num_tokens": 11270351.0,
      "step": 300
    },
    {
      "epoch": 0.23304680038204392,
      "grad_norm": 0.30775734782218933,
      "learning_rate": 4.2169145000427635e-05,
      "loss": 0.1034,
      "mean_token_accuracy": 0.961260712146759,
      "num_tokens": 11454221.0,
      "step": 305
    },
    {
      "epoch": 0.23686723973256923,
      "grad_norm": 24.00685691833496,
      "learning_rate": 4.2137330823191304e-05,
      "loss": 0.1365,
      "mean_token_accuracy": 0.9599539518356324,
      "num_tokens": 11645741.0,
      "step": 310
    },
    {
      "epoch": 0.24068767908309455,
      "grad_norm": 0.31655392050743103,
      "learning_rate": 4.210495279330716e-05,
      "loss": 0.1097,
      "mean_token_accuracy": 0.9635839909315109,
      "num_tokens": 11831492.0,
      "step": 315
    },
    {
      "epoch": 0.24450811843361986,
      "grad_norm": 0.2519204318523407,
      "learning_rate": 4.2072012108177314e-05,
      "loss": 0.1125,
      "mean_token_accuracy": 0.9576877772808075,
      "num_tokens": 12014926.0,
      "step": 320
    },
    {
      "epoch": 0.24832855778414517,
      "grad_norm": 0.32811999320983887,
      "learning_rate": 4.2038509986011984e-05,
      "loss": 0.1029,
      "mean_token_accuracy": 0.9607045829296113,
      "num_tokens": 12207561.0,
      "step": 325
    },
    {
      "epoch": 0.2521489971346705,
      "grad_norm": 0.33064037561416626,
      "learning_rate": 4.200444766578437e-05,
      "loss": 0.0943,
      "mean_token_accuracy": 0.9643021076917648,
      "num_tokens": 12396583.0,
      "step": 330
    },
    {
      "epoch": 0.2559694364851958,
      "grad_norm": 0.16472336649894714,
      "learning_rate": 4.1969826407184916e-05,
      "loss": 0.0971,
      "mean_token_accuracy": 0.9662100911140442,
      "num_tokens": 12579495.0,
      "step": 335
    },
    {
      "epoch": 0.2597898758357211,
      "grad_norm": 0.4175519347190857,
      "learning_rate": 4.1934647490574675e-05,
      "loss": 0.0942,
      "mean_token_accuracy": 0.9644436091184616,
      "num_tokens": 12769617.0,
      "step": 340
    },
    {
      "epoch": 0.2636103151862464,
      "grad_norm": 0.5489321947097778,
      "learning_rate": 4.189891221693798e-05,
      "loss": 0.099,
      "mean_token_accuracy": 0.9652572929859161,
      "num_tokens": 12959397.0,
      "step": 345
    },
    {
      "epoch": 0.26743075453677173,
      "grad_norm": 0.5146623849868774,
      "learning_rate": 4.18626219078343e-05,
      "loss": 0.0845,
      "mean_token_accuracy": 0.9700098544359207,
      "num_tokens": 13148140.0,
      "step": 350
    },
    {
      "epoch": 0.271251193887297,
      "grad_norm": 9.83044147491455,
      "learning_rate": 4.182577790534941e-05,
      "loss": 0.1335,
      "mean_token_accuracy": 0.9588857233524323,
      "num_tokens": 13334926.0,
      "step": 355
    },
    {
      "epoch": 0.27507163323782235,
      "grad_norm": 0.2058829665184021,
      "learning_rate": 4.178838157204572e-05,
      "loss": 0.1033,
      "mean_token_accuracy": 0.9627357661724091,
      "num_tokens": 13523456.0,
      "step": 360
    },
    {
      "epoch": 0.27889207258834764,
      "grad_norm": 0.24188277125358582,
      "learning_rate": 4.1750434290911926e-05,
      "loss": 0.109,
      "mean_token_accuracy": 0.9624377757310867,
      "num_tokens": 13709252.0,
      "step": 365
    },
    {
      "epoch": 0.282712511938873,
      "grad_norm": 0.3566262722015381,
      "learning_rate": 4.171193746531181e-05,
      "loss": 0.0888,
      "mean_token_accuracy": 0.9669180393218995,
      "num_tokens": 13899348.0,
      "step": 370
    },
    {
      "epoch": 0.28653295128939826,
      "grad_norm": 0.3936294913291931,
      "learning_rate": 4.167289251893238e-05,
      "loss": 0.1097,
      "mean_token_accuracy": 0.9597212880849838,
      "num_tokens": 14079661.0,
      "step": 375
    },
    {
      "epoch": 0.2903533906399236,
      "grad_norm": 0.4706442654132843,
      "learning_rate": 4.163330089573122e-05,
      "loss": 0.0985,
      "mean_token_accuracy": 0.9627031296491623,
      "num_tokens": 14261827.0,
      "step": 380
    },
    {
      "epoch": 0.2941738299904489,
      "grad_norm": 0.298458069562912,
      "learning_rate": 4.159316405988308e-05,
      "loss": 0.0964,
      "mean_token_accuracy": 0.9623459368944168,
      "num_tokens": 14450376.0,
      "step": 385
    },
    {
      "epoch": 0.2979942693409742,
      "grad_norm": 0.33347445726394653,
      "learning_rate": 4.15524834957257e-05,
      "loss": 0.1188,
      "mean_token_accuracy": 0.9568544328212738,
      "num_tokens": 14637638.0,
      "step": 390
    },
    {
      "epoch": 0.3018147086914995,
      "grad_norm": 0.32189905643463135,
      "learning_rate": 4.1511260707704986e-05,
      "loss": 0.107,
      "mean_token_accuracy": 0.9608612716197967,
      "num_tokens": 14828104.0,
      "step": 395
    },
    {
      "epoch": 0.30563514804202485,
      "grad_norm": 0.3258993327617645,
      "learning_rate": 4.146949722031929e-05,
      "loss": 0.088,
      "mean_token_accuracy": 0.9665853172540665,
      "num_tokens": 15017072.0,
      "step": 400
    },
    {
      "epoch": 0.30945558739255014,
      "grad_norm": 0.3070586621761322,
      "learning_rate": 4.14271945780631e-05,
      "loss": 0.103,
      "mean_token_accuracy": 0.9601392298936844,
      "num_tokens": 15207268.0,
      "step": 405
    },
    {
      "epoch": 0.3132760267430755,
      "grad_norm": 0.39557355642318726,
      "learning_rate": 4.138435434536988e-05,
      "loss": 0.1213,
      "mean_token_accuracy": 0.9541209667921067,
      "num_tokens": 15394587.0,
      "step": 410
    },
    {
      "epoch": 0.31709646609360076,
      "grad_norm": 0.3308719992637634,
      "learning_rate": 4.1340978106554227e-05,
      "loss": 0.092,
      "mean_token_accuracy": 0.9642503917217254,
      "num_tokens": 15586130.0,
      "step": 415
    },
    {
      "epoch": 0.3209169054441261,
      "grad_norm": 0.20333856344223022,
      "learning_rate": 4.129706746575329e-05,
      "loss": 0.0876,
      "mean_token_accuracy": 0.968433928489685,
      "num_tokens": 15780304.0,
      "step": 420
    },
    {
      "epoch": 0.3247373447946514,
      "grad_norm": 0.3102423846721649,
      "learning_rate": 4.125262404686742e-05,
      "loss": 0.1007,
      "mean_token_accuracy": 0.9613101094961166,
      "num_tokens": 15967608.0,
      "step": 425
    },
    {
      "epoch": 0.32855778414517667,
      "grad_norm": 0.29654639959335327,
      "learning_rate": 4.1207649493500154e-05,
      "loss": 0.0895,
      "mean_token_accuracy": 0.967515516281128,
      "num_tokens": 16157347.0,
      "step": 430
    },
    {
      "epoch": 0.332378223495702,
      "grad_norm": 0.23490051925182343,
      "learning_rate": 4.116214546889738e-05,
      "loss": 0.0975,
      "mean_token_accuracy": 0.9642543792724609,
      "num_tokens": 16346189.0,
      "step": 435
    },
    {
      "epoch": 0.3361986628462273,
      "grad_norm": 0.2675352394580841,
      "learning_rate": 4.111611365588588e-05,
      "loss": 0.0905,
      "mean_token_accuracy": 0.9660478055477142,
      "num_tokens": 16534276.0,
      "step": 440
    },
    {
      "epoch": 0.34001910219675263,
      "grad_norm": 0.25171953439712524,
      "learning_rate": 4.106955575681106e-05,
      "loss": 0.0941,
      "mean_token_accuracy": 0.9624242722988129,
      "num_tokens": 16721141.0,
      "step": 445
    },
    {
      "epoch": 0.3438395415472779,
      "grad_norm": 0.4935978651046753,
      "learning_rate": 4.102247349347401e-05,
      "loss": 0.0827,
      "mean_token_accuracy": 0.9708313107490539,
      "num_tokens": 16912559.0,
      "step": 450
    },
    {
      "epoch": 0.34765998089780326,
      "grad_norm": 0.27168983221054077,
      "learning_rate": 4.097486860706782e-05,
      "loss": 0.0797,
      "mean_token_accuracy": 0.9708745062351227,
      "num_tokens": 17099991.0,
      "step": 455
    },
    {
      "epoch": 0.35148042024832854,
      "grad_norm": 0.3063458204269409,
      "learning_rate": 4.092674285811319e-05,
      "loss": 0.0848,
      "mean_token_accuracy": 0.970548751950264,
      "num_tokens": 17288346.0,
      "step": 460
    },
    {
      "epoch": 0.3553008595988539,
      "grad_norm": 0.3208267390727997,
      "learning_rate": 4.087809802639333e-05,
      "loss": 0.103,
      "mean_token_accuracy": 0.9599024951457977,
      "num_tokens": 17476767.0,
      "step": 465
    },
    {
      "epoch": 0.35912129894937916,
      "grad_norm": 0.4440467059612274,
      "learning_rate": 4.082893591088813e-05,
      "loss": 0.1068,
      "mean_token_accuracy": 0.9598116725683212,
      "num_tokens": 17661898.0,
      "step": 470
    },
    {
      "epoch": 0.3629417382999045,
      "grad_norm": 0.3671613335609436,
      "learning_rate": 4.077925832970764e-05,
      "loss": 0.1008,
      "mean_token_accuracy": 0.963514119386673,
      "num_tokens": 17848891.0,
      "step": 475
    },
    {
      "epoch": 0.3667621776504298,
      "grad_norm": 0.46350836753845215,
      "learning_rate": 4.0729067120024805e-05,
      "loss": 0.0702,
      "mean_token_accuracy": 0.9734578609466553,
      "num_tokens": 18041700.0,
      "step": 480
    },
    {
      "epoch": 0.37058261700095513,
      "grad_norm": 0.23776468634605408,
      "learning_rate": 4.0678364138007574e-05,
      "loss": 0.0941,
      "mean_token_accuracy": 0.9612627565860749,
      "num_tokens": 18232213.0,
      "step": 485
    },
    {
      "epoch": 0.3744030563514804,
      "grad_norm": 0.25439709424972534,
      "learning_rate": 4.062715125875021e-05,
      "loss": 0.089,
      "mean_token_accuracy": 0.9673547148704529,
      "num_tokens": 18417525.0,
      "step": 490
    },
    {
      "epoch": 0.37822349570200575,
      "grad_norm": 0.23942908644676208,
      "learning_rate": 4.0575430376203964e-05,
      "loss": 0.0875,
      "mean_token_accuracy": 0.9648292303085327,
      "num_tokens": 18611286.0,
      "step": 495
    },
    {
      "epoch": 0.38204393505253104,
      "grad_norm": 0.4034489095211029,
      "learning_rate": 4.0523203403107025e-05,
      "loss": 0.0884,
      "mean_token_accuracy": 0.9656334936618804,
      "num_tokens": 18805075.0,
      "step": 500
    },
    {
      "epoch": 0.3858643744030564,
      "grad_norm": 0.3862541615962982,
      "learning_rate": 4.0470472270913807e-05,
      "loss": 0.0944,
      "mean_token_accuracy": 0.9657697170972824,
      "num_tokens": 18994710.0,
      "step": 505
    },
    {
      "epoch": 0.38968481375358166,
      "grad_norm": 0.2153453230857849,
      "learning_rate": 4.041723892972348e-05,
      "loss": 0.0999,
      "mean_token_accuracy": 0.9615686565637589,
      "num_tokens": 19182207.0,
      "step": 510
    },
    {
      "epoch": 0.39350525310410694,
      "grad_norm": 0.26945939660072327,
      "learning_rate": 4.036350534820788e-05,
      "loss": 0.0815,
      "mean_token_accuracy": 0.9687103390693664,
      "num_tokens": 19371393.0,
      "step": 515
    },
    {
      "epoch": 0.3973256924546323,
      "grad_norm": 0.3088325262069702,
      "learning_rate": 4.0309273513538705e-05,
      "loss": 0.0842,
      "mean_token_accuracy": 0.9696242094039917,
      "num_tokens": 19556798.0,
      "step": 520
    },
    {
      "epoch": 0.40114613180515757,
      "grad_norm": 0.3315938711166382,
      "learning_rate": 4.025454543131401e-05,
      "loss": 0.1068,
      "mean_token_accuracy": 0.9622175186872483,
      "num_tokens": 19744574.0,
      "step": 525
    },
    {
      "epoch": 0.4049665711556829,
      "grad_norm": 0.2412356436252594,
      "learning_rate": 4.019932312548405e-05,
      "loss": 0.0985,
      "mean_token_accuracy": 0.9639840573072433,
      "num_tokens": 19927652.0,
      "step": 530
    },
    {
      "epoch": 0.4087870105062082,
      "grad_norm": 0.35652726888656616,
      "learning_rate": 4.0143608638276414e-05,
      "loss": 0.1005,
      "mean_token_accuracy": 0.9605095684528351,
      "num_tokens": 20115704.0,
      "step": 535
    },
    {
      "epoch": 0.41260744985673353,
      "grad_norm": 0.29808974266052246,
      "learning_rate": 4.008740403012052e-05,
      "loss": 0.0861,
      "mean_token_accuracy": 0.9695513725280762,
      "num_tokens": 20305414.0,
      "step": 540
    },
    {
      "epoch": 0.4164278892072588,
      "grad_norm": 0.36241698265075684,
      "learning_rate": 4.0030711379571394e-05,
      "loss": 0.0931,
      "mean_token_accuracy": 0.9662731915712357,
      "num_tokens": 20488362.0,
      "step": 545
    },
    {
      "epoch": 0.42024832855778416,
      "grad_norm": 0.2920278012752533,
      "learning_rate": 3.997353278323283e-05,
      "loss": 0.0823,
      "mean_token_accuracy": 0.9721543341875076,
      "num_tokens": 20676907.0,
      "step": 550
    },
    {
      "epoch": 0.42406876790830944,
      "grad_norm": 0.40454885363578796,
      "learning_rate": 3.9915870355679814e-05,
      "loss": 0.0891,
      "mean_token_accuracy": 0.9684916347265243,
      "num_tokens": 20865482.0,
      "step": 555
    },
    {
      "epoch": 0.4278892072588348,
      "grad_norm": 0.2907525599002838,
      "learning_rate": 3.985772622938037e-05,
      "loss": 0.091,
      "mean_token_accuracy": 0.9653962075710296,
      "num_tokens": 21049226.0,
      "step": 560
    },
    {
      "epoch": 0.43170964660936006,
      "grad_norm": 0.28567659854888916,
      "learning_rate": 3.979910255461665e-05,
      "loss": 0.0853,
      "mean_token_accuracy": 0.965234687924385,
      "num_tokens": 21233739.0,
      "step": 565
    },
    {
      "epoch": 0.4355300859598854,
      "grad_norm": 0.1904965341091156,
      "learning_rate": 3.9740001499405454e-05,
      "loss": 0.0853,
      "mean_token_accuracy": 0.967443099617958,
      "num_tokens": 21422679.0,
      "step": 570
    },
    {
      "epoch": 0.4393505253104107,
      "grad_norm": 0.30602169036865234,
      "learning_rate": 3.968042524941802e-05,
      "loss": 0.0974,
      "mean_token_accuracy": 0.9634604036808014,
      "num_tokens": 21613600.0,
      "step": 575
    },
    {
      "epoch": 0.44317096466093603,
      "grad_norm": 0.1461438238620758,
      "learning_rate": 3.9620376007899225e-05,
      "loss": 0.0823,
      "mean_token_accuracy": 0.9677056491374969,
      "num_tokens": 21803684.0,
      "step": 580
    },
    {
      "epoch": 0.4469914040114613,
      "grad_norm": 0.2723187506198883,
      "learning_rate": 3.955985599558606e-05,
      "loss": 0.0777,
      "mean_token_accuracy": 0.9715256482362747,
      "num_tokens": 21993256.0,
      "step": 585
    },
    {
      "epoch": 0.45081184336198665,
      "grad_norm": 0.31419074535369873,
      "learning_rate": 3.949886745062554e-05,
      "loss": 0.0923,
      "mean_token_accuracy": 0.967405590415001,
      "num_tokens": 22181316.0,
      "step": 590
    },
    {
      "epoch": 0.45463228271251194,
      "grad_norm": 0.2928747534751892,
      "learning_rate": 3.943741262849194e-05,
      "loss": 0.1068,
      "mean_token_accuracy": 0.9629636943340302,
      "num_tokens": 22368006.0,
      "step": 595
    },
    {
      "epoch": 0.4584527220630373,
      "grad_norm": 0.35084590315818787,
      "learning_rate": 3.937549380190334e-05,
      "loss": 0.0839,
      "mean_token_accuracy": 0.9666139155626297,
      "num_tokens": 22555541.0,
      "step": 600
    },
    {
      "epoch": 0.46227316141356256,
      "grad_norm": 0.4395831227302551,
      "learning_rate": 3.931311326073763e-05,
      "loss": 0.102,
      "mean_token_accuracy": 0.9619550377130508,
      "num_tokens": 22749056.0,
      "step": 605
    },
    {
      "epoch": 0.46609360076408785,
      "grad_norm": 0.24934665858745575,
      "learning_rate": 3.9250273311947766e-05,
      "loss": 0.0954,
      "mean_token_accuracy": 0.967662912607193,
      "num_tokens": 22938308.0,
      "step": 610
    },
    {
      "epoch": 0.4699140401146132,
      "grad_norm": 0.29533061385154724,
      "learning_rate": 3.918697627947652e-05,
      "loss": 0.0741,
      "mean_token_accuracy": 0.9715071082115173,
      "num_tokens": 23122307.0,
      "step": 615
    },
    {
      "epoch": 0.47373447946513847,
      "grad_norm": 0.21731697022914886,
      "learning_rate": 3.912322450417046e-05,
      "loss": 0.0779,
      "mean_token_accuracy": 0.9680689096450805,
      "num_tokens": 23308793.0,
      "step": 620
    },
    {
      "epoch": 0.4775549188156638,
      "grad_norm": 0.1922738403081894,
      "learning_rate": 3.905902034369347e-05,
      "loss": 0.1033,
      "mean_token_accuracy": 0.9615610003471374,
      "num_tokens": 23496680.0,
      "step": 625
    },
    {
      "epoch": 0.4813753581661891,
      "grad_norm": 0.17477506399154663,
      "learning_rate": 3.899436617243947e-05,
      "loss": 0.0982,
      "mean_token_accuracy": 0.9618253350257874,
      "num_tokens": 23682970.0,
      "step": 630
    },
    {
      "epoch": 0.48519579751671443,
      "grad_norm": 1.180742621421814,
      "learning_rate": 3.892926438144468e-05,
      "loss": 0.0876,
      "mean_token_accuracy": 0.9670244246721268,
      "num_tokens": 23871273.0,
      "step": 635
    },
    {
      "epoch": 0.4890162368672397,
      "grad_norm": 0.25482797622680664,
      "learning_rate": 3.886371737829914e-05,
      "loss": 0.0846,
      "mean_token_accuracy": 0.9713111281394958,
      "num_tokens": 24059802.0,
      "step": 640
    },
    {
      "epoch": 0.49283667621776506,
      "grad_norm": 0.21185167133808136,
      "learning_rate": 3.8797727587057716e-05,
      "loss": 0.0953,
      "mean_token_accuracy": 0.9666842132806778,
      "num_tokens": 24249407.0,
      "step": 645
    },
    {
      "epoch": 0.49665711556829034,
      "grad_norm": 0.4857417345046997,
      "learning_rate": 3.8731297448150406e-05,
      "loss": 0.0873,
      "mean_token_accuracy": 0.9672016590833664,
      "num_tokens": 24438420.0,
      "step": 650
    },
    {
      "epoch": 0.5004775549188156,
      "grad_norm": 0.2848888635635376,
      "learning_rate": 3.8664429418292145e-05,
      "loss": 0.0912,
      "mean_token_accuracy": 0.9696382105350494,
      "num_tokens": 24626478.0,
      "step": 655
    },
    {
      "epoch": 0.504297994269341,
      "grad_norm": 0.24754320085048676,
      "learning_rate": 3.859712597039192e-05,
      "loss": 0.0853,
      "mean_token_accuracy": 0.9701752871274948,
      "num_tokens": 24816459.0,
      "step": 660
    },
    {
      "epoch": 0.5081184336198663,
      "grad_norm": 0.25339099764823914,
      "learning_rate": 3.852938959346128e-05,
      "loss": 0.1,
      "mean_token_accuracy": 0.964874091744423,
      "num_tokens": 25004787.0,
      "step": 665
    },
    {
      "epoch": 0.5119388729703916,
      "grad_norm": 0.2692340016365051,
      "learning_rate": 3.846122279252237e-05,
      "loss": 0.072,
      "mean_token_accuracy": 0.9739194989204407,
      "num_tokens": 25194290.0,
      "step": 670
    },
    {
      "epoch": 0.5157593123209169,
      "grad_norm": 0.29502665996551514,
      "learning_rate": 3.839262808851524e-05,
      "loss": 0.0983,
      "mean_token_accuracy": 0.9636133432388305,
      "num_tokens": 25377670.0,
      "step": 675
    },
    {
      "epoch": 0.5195797516714422,
      "grad_norm": 0.20753639936447144,
      "learning_rate": 3.832360801820459e-05,
      "loss": 0.0683,
      "mean_token_accuracy": 0.9763543248176575,
      "num_tokens": 25570694.0,
      "step": 680
    },
    {
      "epoch": 0.5234001910219676,
      "grad_norm": 0.33412817120552063,
      "learning_rate": 3.825416513408604e-05,
      "loss": 0.0889,
      "mean_token_accuracy": 0.9706046104431152,
      "num_tokens": 25761296.0,
      "step": 685
    },
    {
      "epoch": 0.5272206303724928,
      "grad_norm": 0.21740436553955078,
      "learning_rate": 3.818430200429163e-05,
      "loss": 0.1045,
      "mean_token_accuracy": 0.9615556567907333,
      "num_tokens": 25948850.0,
      "step": 690
    },
    {
      "epoch": 0.5310410697230181,
      "grad_norm": 0.19813643395900726,
      "learning_rate": 3.811402121249493e-05,
      "loss": 0.093,
      "mean_token_accuracy": 0.9672826796770095,
      "num_tokens": 26139523.0,
      "step": 695
    },
    {
      "epoch": 0.5348615090735435,
      "grad_norm": 0.3105727434158325,
      "learning_rate": 3.8043325357815445e-05,
      "loss": 0.0695,
      "mean_token_accuracy": 0.9759637385606765,
      "num_tokens": 26335435.0,
      "step": 700
    },
    {
      "epoch": 0.5386819484240688,
      "grad_norm": 0.1332012563943863,
      "learning_rate": 3.797221705472252e-05,
      "loss": 0.0805,
      "mean_token_accuracy": 0.9682069540023803,
      "num_tokens": 26526942.0,
      "step": 705
    },
    {
      "epoch": 0.542502387774594,
      "grad_norm": 0.30462008714675903,
      "learning_rate": 3.790069893293862e-05,
      "loss": 0.0953,
      "mean_token_accuracy": 0.9650699675083161,
      "num_tokens": 26713213.0,
      "step": 710
    },
    {
      "epoch": 0.5463228271251194,
      "grad_norm": 0.17359836399555206,
      "learning_rate": 3.7828773637342113e-05,
      "loss": 0.0946,
      "mean_token_accuracy": 0.9641090750694274,
      "num_tokens": 26901468.0,
      "step": 715
    },
    {
      "epoch": 0.5501432664756447,
      "grad_norm": 0.19708728790283203,
      "learning_rate": 3.775644382786943e-05,
      "loss": 0.079,
      "mean_token_accuracy": 0.969389620423317,
      "num_tokens": 27092825.0,
      "step": 720
    },
    {
      "epoch": 0.55396370582617,
      "grad_norm": 0.29853788018226624,
      "learning_rate": 3.768371217941672e-05,
      "loss": 0.0862,
      "mean_token_accuracy": 0.9652405500411987,
      "num_tokens": 27282154.0,
      "step": 725
    },
    {
      "epoch": 0.5577841451766953,
      "grad_norm": 0.12112744152545929,
      "learning_rate": 3.761058138174091e-05,
      "loss": 0.0705,
      "mean_token_accuracy": 0.9710530459880828,
      "num_tokens": 27474198.0,
      "step": 730
    },
    {
      "epoch": 0.5616045845272206,
      "grad_norm": 0.24082429707050323,
      "learning_rate": 3.753705413936023e-05,
      "loss": 0.0926,
      "mean_token_accuracy": 0.9674068808555603,
      "num_tokens": 27659478.0,
      "step": 735
    },
    {
      "epoch": 0.565425023877746,
      "grad_norm": 0.25327593088150024,
      "learning_rate": 3.746313317145419e-05,
      "loss": 0.0955,
      "mean_token_accuracy": 0.963821080327034,
      "num_tokens": 27841586.0,
      "step": 740
    },
    {
      "epoch": 0.5692454632282713,
      "grad_norm": 0.3777076303958893,
      "learning_rate": 3.7388821211763076e-05,
      "loss": 0.0854,
      "mean_token_accuracy": 0.9695561289787292,
      "num_tokens": 28030164.0,
      "step": 745
    },
    {
      "epoch": 0.5730659025787965,
      "grad_norm": 0.23385760188102722,
      "learning_rate": 3.7314121008486755e-05,
      "loss": 0.08,
      "mean_token_accuracy": 0.9686681360006333,
      "num_tokens": 28221554.0,
      "step": 750
    },
    {
      "epoch": 0.5768863419293219,
      "grad_norm": 0.22515222430229187,
      "learning_rate": 3.723903532418312e-05,
      "loss": 0.095,
      "mean_token_accuracy": 0.9663045287132264,
      "num_tokens": 28405780.0,
      "step": 755
    },
    {
      "epoch": 0.5807067812798472,
      "grad_norm": 0.23947760462760925,
      "learning_rate": 3.716356693566589e-05,
      "loss": 0.1043,
      "mean_token_accuracy": 0.9601571738719941,
      "num_tokens": 28589164.0,
      "step": 760
    },
    {
      "epoch": 0.5845272206303725,
      "grad_norm": 0.21670001745224,
      "learning_rate": 3.708771863390193e-05,
      "loss": 0.0855,
      "mean_token_accuracy": 0.9669242978096009,
      "num_tokens": 28782866.0,
      "step": 765
    },
    {
      "epoch": 0.5883476599808978,
      "grad_norm": 0.27755483984947205,
      "learning_rate": 3.701149322390802e-05,
      "loss": 0.0769,
      "mean_token_accuracy": 0.9709088444709778,
      "num_tokens": 28973298.0,
      "step": 770
    },
    {
      "epoch": 0.5921680993314231,
      "grad_norm": 0.3821544945240021,
      "learning_rate": 3.693489352464713e-05,
      "loss": 0.0864,
      "mean_token_accuracy": 0.9679794996976853,
      "num_tokens": 29164313.0,
      "step": 775
    },
    {
      "epoch": 0.5959885386819485,
      "grad_norm": 0.2872138023376465,
      "learning_rate": 3.68579223689242e-05,
      "loss": 0.096,
      "mean_token_accuracy": 0.9626093208789825,
      "num_tokens": 29348960.0,
      "step": 780
    },
    {
      "epoch": 0.5998089780324737,
      "grad_norm": 0.4298568069934845,
      "learning_rate": 3.678058260328131e-05,
      "loss": 0.081,
      "mean_token_accuracy": 0.9725433737039566,
      "num_tokens": 29539282.0,
      "step": 785
    },
    {
      "epoch": 0.603629417382999,
      "grad_norm": 0.3037995994091034,
      "learning_rate": 3.670287708789248e-05,
      "loss": 0.0892,
      "mean_token_accuracy": 0.9649462342262268,
      "num_tokens": 29724820.0,
      "step": 790
    },
    {
      "epoch": 0.6074498567335244,
      "grad_norm": 0.3658304214477539,
      "learning_rate": 3.6624808696457835e-05,
      "loss": 0.1042,
      "mean_token_accuracy": 0.9630384862422943,
      "num_tokens": 29906584.0,
      "step": 795
    },
    {
      "epoch": 0.6112702960840497,
      "grad_norm": 0.32044458389282227,
      "learning_rate": 3.6546380316097366e-05,
      "loss": 0.0846,
      "mean_token_accuracy": 0.9699332922697067,
      "num_tokens": 30099681.0,
      "step": 800
    },
    {
      "epoch": 0.6150907354345749,
      "grad_norm": 0.3886227309703827,
      "learning_rate": 3.646759484724415e-05,
      "loss": 0.0887,
      "mean_token_accuracy": 0.9664107114076614,
      "num_tokens": 30288264.0,
      "step": 805
    },
    {
      "epoch": 0.6189111747851003,
      "grad_norm": 0.24491894245147705,
      "learning_rate": 3.638845520353711e-05,
      "loss": 0.0762,
      "mean_token_accuracy": 0.9727777987718582,
      "num_tokens": 30474509.0,
      "step": 810
    },
    {
      "epoch": 0.6227316141356256,
      "grad_norm": 0.22406314313411713,
      "learning_rate": 3.6308964311713196e-05,
      "loss": 0.0747,
      "mean_token_accuracy": 0.971963056921959,
      "num_tokens": 30672948.0,
      "step": 815
    },
    {
      "epoch": 0.626552053486151,
      "grad_norm": 0.23877441883087158,
      "learning_rate": 3.6229125111499226e-05,
      "loss": 0.0851,
      "mean_token_accuracy": 0.9684547692537308,
      "num_tokens": 30868012.0,
      "step": 820
    },
    {
      "epoch": 0.6303724928366762,
      "grad_norm": 0.19574357569217682,
      "learning_rate": 3.6148940555503124e-05,
      "loss": 0.0931,
      "mean_token_accuracy": 0.9639934927225113,
      "num_tokens": 31058865.0,
      "step": 825
    },
    {
      "epoch": 0.6341929321872015,
      "grad_norm": 0.2133413553237915,
      "learning_rate": 3.6068413609104744e-05,
      "loss": 0.1011,
      "mean_token_accuracy": 0.9611745178699493,
      "num_tokens": 31239751.0,
      "step": 830
    },
    {
      "epoch": 0.6380133715377269,
      "grad_norm": 0.16554482281208038,
      "learning_rate": 3.598754725034619e-05,
      "loss": 0.0762,
      "mean_token_accuracy": 0.9709636360406876,
      "num_tokens": 31420307.0,
      "step": 835
    },
    {
      "epoch": 0.6418338108882522,
      "grad_norm": 0.23217220604419708,
      "learning_rate": 3.5906344469821684e-05,
      "loss": 0.0861,
      "mean_token_accuracy": 0.9689477831125259,
      "num_tokens": 31606840.0,
      "step": 840
    },
    {
      "epoch": 0.6456542502387774,
      "grad_norm": 0.2894331216812134,
      "learning_rate": 3.5824808270567e-05,
      "loss": 0.0953,
      "mean_token_accuracy": 0.9652222245931625,
      "num_tokens": 31795330.0,
      "step": 845
    },
    {
      "epoch": 0.6494746895893028,
      "grad_norm": 0.15870925784111023,
      "learning_rate": 3.574294166794837e-05,
      "loss": 0.0688,
      "mean_token_accuracy": 0.9743613868951797,
      "num_tokens": 31980238.0,
      "step": 850
    },
    {
      "epoch": 0.6532951289398281,
      "grad_norm": 0.38253387808799744,
      "learning_rate": 3.566074768955096e-05,
      "loss": 0.0895,
      "mean_token_accuracy": 0.9676443010568618,
      "num_tokens": 32168498.0,
      "step": 855
    },
    {
      "epoch": 0.6571155682903533,
      "grad_norm": 0.24848639965057373,
      "learning_rate": 3.557822937506696e-05,
      "loss": 0.0858,
      "mean_token_accuracy": 0.9665856540203095,
      "num_tokens": 32354931.0,
      "step": 860
    },
    {
      "epoch": 0.6609360076408787,
      "grad_norm": 0.3941405117511749,
      "learning_rate": 3.549538977618311e-05,
      "loss": 0.1056,
      "mean_token_accuracy": 0.9621305525302887,
      "num_tokens": 32538801.0,
      "step": 865
    },
    {
      "epoch": 0.664756446991404,
      "grad_norm": 0.26622191071510315,
      "learning_rate": 3.5412231956467874e-05,
      "loss": 0.0992,
      "mean_token_accuracy": 0.9625103205442429,
      "num_tokens": 32728446.0,
      "step": 870
    },
    {
      "epoch": 0.6685768863419294,
      "grad_norm": 0.1775600016117096,
      "learning_rate": 3.5328758991258146e-05,
      "loss": 0.091,
      "mean_token_accuracy": 0.9657710462808609,
      "num_tokens": 32917934.0,
      "step": 875
    },
    {
      "epoch": 0.6723973256924546,
      "grad_norm": 0.2797180414199829,
      "learning_rate": 3.524497396754551e-05,
      "loss": 0.0958,
      "mean_token_accuracy": 0.9660986751317978,
      "num_tokens": 33107410.0,
      "step": 880
    },
    {
      "epoch": 0.6762177650429799,
      "grad_norm": 0.21887938678264618,
      "learning_rate": 3.5160879983862046e-05,
      "loss": 0.0837,
      "mean_token_accuracy": 0.9694306135177613,
      "num_tokens": 33297869.0,
      "step": 885
    },
    {
      "epoch": 0.6800382043935053,
      "grad_norm": 0.2001357227563858,
      "learning_rate": 3.5076480150165817e-05,
      "loss": 0.0714,
      "mean_token_accuracy": 0.9740285396575927,
      "num_tokens": 33493898.0,
      "step": 890
    },
    {
      "epoch": 0.6838586437440306,
      "grad_norm": 0.23507362604141235,
      "learning_rate": 3.4991777587725774e-05,
      "loss": 0.0929,
      "mean_token_accuracy": 0.965422710776329,
      "num_tokens": 33678191.0,
      "step": 895
    },
    {
      "epoch": 0.6876790830945558,
      "grad_norm": 0.3677293062210083,
      "learning_rate": 3.4906775429006386e-05,
      "loss": 0.0743,
      "mean_token_accuracy": 0.9716667503118515,
      "num_tokens": 33871419.0,
      "step": 900
    },
    {
      "epoch": 0.6914995224450812,
      "grad_norm": 0.16620738804340363,
      "learning_rate": 3.4821476817551755e-05,
      "loss": 0.0915,
      "mean_token_accuracy": 0.9655976742506027,
      "num_tokens": 34055370.0,
      "step": 905
    },
    {
      "epoch": 0.6953199617956065,
      "grad_norm": 0.3682331144809723,
      "learning_rate": 3.473588490786939e-05,
      "loss": 0.081,
      "mean_token_accuracy": 0.9663502663373947,
      "num_tokens": 34247370.0,
      "step": 910
    },
    {
      "epoch": 0.6991404011461319,
      "grad_norm": 0.18665063381195068,
      "learning_rate": 3.465000286531353e-05,
      "loss": 0.0784,
      "mean_token_accuracy": 0.9698567777872086,
      "num_tokens": 34437248.0,
      "step": 915
    },
    {
      "epoch": 0.7029608404966571,
      "grad_norm": 0.26217788457870483,
      "learning_rate": 3.456383386596808e-05,
      "loss": 0.0787,
      "mean_token_accuracy": 0.9704749584197998,
      "num_tokens": 34628771.0,
      "step": 920
    },
    {
      "epoch": 0.7067812798471824,
      "grad_norm": 0.285917192697525,
      "learning_rate": 3.447738109652918e-05,
      "loss": 0.0902,
      "mean_token_accuracy": 0.9661570072174073,
      "num_tokens": 34818100.0,
      "step": 925
    },
    {
      "epoch": 0.7106017191977078,
      "grad_norm": 0.2116532325744629,
      "learning_rate": 3.439064775418731e-05,
      "loss": 0.0853,
      "mean_token_accuracy": 0.9665609240531922,
      "num_tokens": 35011505.0,
      "step": 930
    },
    {
      "epoch": 0.7144221585482331,
      "grad_norm": 0.21706698834896088,
      "learning_rate": 3.430363704650912e-05,
      "loss": 0.0947,
      "mean_token_accuracy": 0.9647042155265808,
      "num_tokens": 35199923.0,
      "step": 935
    },
    {
      "epoch": 0.7182425978987583,
      "grad_norm": 0.16597236692905426,
      "learning_rate": 3.421635219131872e-05,
      "loss": 0.0698,
      "mean_token_accuracy": 0.9757835417985916,
      "num_tokens": 35389697.0,
      "step": 940
    },
    {
      "epoch": 0.7220630372492837,
      "grad_norm": 0.2854805886745453,
      "learning_rate": 3.412879641657875e-05,
      "loss": 0.0919,
      "mean_token_accuracy": 0.9653942167758942,
      "num_tokens": 35575373.0,
      "step": 945
    },
    {
      "epoch": 0.725883476599809,
      "grad_norm": 0.14670076966285706,
      "learning_rate": 3.404097296027098e-05,
      "loss": 0.0796,
      "mean_token_accuracy": 0.9704311460256576,
      "num_tokens": 35765389.0,
      "step": 950
    },
    {
      "epoch": 0.7297039159503342,
      "grad_norm": 0.22133147716522217,
      "learning_rate": 3.395288507027655e-05,
      "loss": 0.0862,
      "mean_token_accuracy": 0.9677942723035813,
      "num_tokens": 35953675.0,
      "step": 955
    },
    {
      "epoch": 0.7335243553008596,
      "grad_norm": 0.19771571457386017,
      "learning_rate": 3.386453600425588e-05,
      "loss": 0.0979,
      "mean_token_accuracy": 0.9636722028255462,
      "num_tokens": 36143681.0,
      "step": 960
    },
    {
      "epoch": 0.7373447946513849,
      "grad_norm": 0.21256014704704285,
      "learning_rate": 3.3775929029528174e-05,
      "loss": 0.0705,
      "mean_token_accuracy": 0.9728881269693375,
      "num_tokens": 36334495.0,
      "step": 965
    },
    {
      "epoch": 0.7411652340019103,
      "grad_norm": 0.23361259698867798,
      "learning_rate": 3.368706742295061e-05,
      "loss": 0.0689,
      "mean_token_accuracy": 0.9734117150306701,
      "num_tokens": 36523096.0,
      "step": 970
    },
    {
      "epoch": 0.7449856733524355,
      "grad_norm": 0.12377287447452545,
      "learning_rate": 3.359795447079712e-05,
      "loss": 0.1063,
      "mean_token_accuracy": 0.9605233788490295,
      "num_tokens": 36705458.0,
      "step": 975
    },
    {
      "epoch": 0.7488061127029608,
      "grad_norm": 0.20916534960269928,
      "learning_rate": 3.3508593468636916e-05,
      "loss": 0.0859,
      "mean_token_accuracy": 0.9657229632139206,
      "num_tokens": 36893580.0,
      "step": 980
    },
    {
      "epoch": 0.7526265520534862,
      "grad_norm": 0.29062730073928833,
      "learning_rate": 3.341898772121256e-05,
      "loss": 0.0652,
      "mean_token_accuracy": 0.9765775531530381,
      "num_tokens": 37081201.0,
      "step": 985
    },
    {
      "epoch": 0.7564469914040115,
      "grad_norm": 0.12679129838943481,
      "learning_rate": 3.3329140542317766e-05,
      "loss": 0.0694,
      "mean_token_accuracy": 0.9743989706039429,
      "num_tokens": 37271871.0,
      "step": 990
    },
    {
      "epoch": 0.7602674307545367,
      "grad_norm": 0.2163613736629486,
      "learning_rate": 3.3239055254674843e-05,
      "loss": 0.0848,
      "mean_token_accuracy": 0.9658847957849502,
      "num_tokens": 37459173.0,
      "step": 995
    },
    {
      "epoch": 0.7640878701050621,
      "grad_norm": 0.2529585063457489,
      "learning_rate": 3.314873518981184e-05,
      "loss": 0.0785,
      "mean_token_accuracy": 0.9713763266801834,
      "num_tokens": 37648824.0,
      "step": 1000
    }
  ],
  "logging_steps": 5,
  "max_steps": 2618,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.07112036183441e+18,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}