{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9952499434517078,
  "eval_steps": 500,
  "global_step": 4400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004523863379325944,
      "grad_norm": 1.2934283018112183,
      "learning_rate": 1.1299435028248588e-07,
      "loss": 1.8133,
      "step": 20
    },
    {
      "epoch": 0.009047726758651889,
      "grad_norm": 1.8428480625152588,
      "learning_rate": 2.2598870056497177e-07,
      "loss": 1.6188,
      "step": 40
    },
    {
      "epoch": 0.013571590137977834,
      "grad_norm": 1.6636091470718384,
      "learning_rate": 3.3898305084745766e-07,
      "loss": 1.6899,
      "step": 60
    },
    {
      "epoch": 0.018095453517303777,
      "grad_norm": 2.0820469856262207,
      "learning_rate": 4.5197740112994353e-07,
      "loss": 1.703,
      "step": 80
    },
    {
      "epoch": 0.022619316896629722,
      "grad_norm": 2.4241092205047607,
      "learning_rate": 5.649717514124295e-07,
      "loss": 1.7203,
      "step": 100
    },
    {
      "epoch": 0.027143180275955667,
      "grad_norm": 4.02646017074585,
      "learning_rate": 6.779661016949153e-07,
      "loss": 1.6147,
      "step": 120
    },
    {
      "epoch": 0.03166704365528161,
      "grad_norm": 2.3496134281158447,
      "learning_rate": 7.909604519774013e-07,
      "loss": 1.6992,
      "step": 140
    },
    {
      "epoch": 0.036190907034607554,
      "grad_norm": 2.2215747833251953,
      "learning_rate": 9.039548022598871e-07,
      "loss": 1.4976,
      "step": 160
    },
    {
      "epoch": 0.040714770413933496,
      "grad_norm": 4.31510066986084,
      "learning_rate": 1.016949152542373e-06,
      "loss": 1.4859,
      "step": 180
    },
    {
      "epoch": 0.045238633793259445,
      "grad_norm": 2.950899600982666,
      "learning_rate": 1.129943502824859e-06,
      "loss": 1.3391,
      "step": 200
    },
    {
      "epoch": 0.049762497172585386,
      "grad_norm": 4.141122341156006,
      "learning_rate": 1.2429378531073449e-06,
      "loss": 1.3377,
      "step": 220
    },
    {
      "epoch": 0.054286360551911335,
      "grad_norm": 2.346282720565796,
      "learning_rate": 1.3559322033898307e-06,
      "loss": 1.2225,
      "step": 240
    },
    {
      "epoch": 0.05881022393123728,
      "grad_norm": 1.5768523216247559,
      "learning_rate": 1.4689265536723166e-06,
      "loss": 1.1489,
      "step": 260
    },
    {
      "epoch": 0.06333408731056323,
      "grad_norm": 1.6714472770690918,
      "learning_rate": 1.5819209039548026e-06,
      "loss": 1.2574,
      "step": 280
    },
    {
      "epoch": 0.06785795068988916,
      "grad_norm": 1.2765991687774658,
      "learning_rate": 1.6949152542372882e-06,
      "loss": 1.0654,
      "step": 300
    },
    {
      "epoch": 0.07238181406921511,
      "grad_norm": 0.9537855386734009,
      "learning_rate": 1.8079096045197741e-06,
      "loss": 1.0468,
      "step": 320
    },
    {
      "epoch": 0.07690567744854106,
      "grad_norm": 0.8233529329299927,
      "learning_rate": 1.92090395480226e-06,
      "loss": 1.0045,
      "step": 340
    },
    {
      "epoch": 0.08142954082786699,
      "grad_norm": 0.977318286895752,
      "learning_rate": 2.033898305084746e-06,
      "loss": 1.0425,
      "step": 360
    },
    {
      "epoch": 0.08595340420719294,
      "grad_norm": 0.7735580205917358,
      "learning_rate": 2.146892655367232e-06,
      "loss": 0.8498,
      "step": 380
    },
    {
      "epoch": 0.09047726758651889,
      "grad_norm": 0.9134358763694763,
      "learning_rate": 2.259887005649718e-06,
      "loss": 0.8149,
      "step": 400
    },
    {
      "epoch": 0.09500113096584484,
      "grad_norm": 1.034114956855774,
      "learning_rate": 2.372881355932204e-06,
      "loss": 0.9009,
      "step": 420
    },
    {
      "epoch": 0.09952499434517077,
      "grad_norm": 0.8178341388702393,
      "learning_rate": 2.4858757062146898e-06,
      "loss": 0.8489,
      "step": 440
    },
    {
      "epoch": 0.10404885772449672,
      "grad_norm": 0.9051856398582458,
      "learning_rate": 2.5988700564971753e-06,
      "loss": 0.8824,
      "step": 460
    },
    {
      "epoch": 0.10857272110382267,
      "grad_norm": 0.7567653059959412,
      "learning_rate": 2.7118644067796613e-06,
      "loss": 0.8307,
      "step": 480
    },
    {
      "epoch": 0.1130965844831486,
      "grad_norm": 0.800140917301178,
      "learning_rate": 2.8248587570621473e-06,
      "loss": 0.7441,
      "step": 500
    },
    {
      "epoch": 0.11762044786247455,
      "grad_norm": 0.6388680338859558,
      "learning_rate": 2.9378531073446333e-06,
      "loss": 0.7953,
      "step": 520
    },
    {
      "epoch": 0.1221443112418005,
      "grad_norm": 1.6128122806549072,
      "learning_rate": 3.0508474576271192e-06,
      "loss": 0.7497,
      "step": 540
    },
    {
      "epoch": 0.12666817462112645,
      "grad_norm": 0.7322567105293274,
      "learning_rate": 3.163841807909605e-06,
      "loss": 0.7005,
      "step": 560
    },
    {
      "epoch": 0.1311920380004524,
      "grad_norm": 0.678276777267456,
      "learning_rate": 3.2768361581920903e-06,
      "loss": 0.678,
      "step": 580
    },
    {
      "epoch": 0.13571590137977832,
      "grad_norm": 0.5271691679954529,
      "learning_rate": 3.3898305084745763e-06,
      "loss": 0.6971,
      "step": 600
    },
    {
      "epoch": 0.14023976475910427,
      "grad_norm": 1.0406115055084229,
      "learning_rate": 3.5028248587570623e-06,
      "loss": 0.6432,
      "step": 620
    },
    {
      "epoch": 0.14476362813843022,
      "grad_norm": 1.0237056016921997,
      "learning_rate": 3.6158192090395483e-06,
      "loss": 0.6278,
      "step": 640
    },
    {
      "epoch": 0.14928749151775617,
      "grad_norm": 0.4786810278892517,
      "learning_rate": 3.7288135593220342e-06,
      "loss": 0.6545,
      "step": 660
    },
    {
      "epoch": 0.15381135489708211,
      "grad_norm": 0.9167463183403015,
      "learning_rate": 3.84180790960452e-06,
      "loss": 0.5941,
      "step": 680
    },
    {
      "epoch": 0.15833521827640806,
      "grad_norm": 0.764803409576416,
      "learning_rate": 3.954802259887006e-06,
      "loss": 0.5993,
      "step": 700
    },
    {
      "epoch": 0.16285908165573398,
      "grad_norm": 0.5060482025146484,
      "learning_rate": 4.067796610169492e-06,
      "loss": 0.7018,
      "step": 720
    },
    {
      "epoch": 0.16738294503505993,
      "grad_norm": 1.2368682622909546,
      "learning_rate": 4.180790960451978e-06,
      "loss": 0.656,
      "step": 740
    },
    {
      "epoch": 0.17190680841438588,
      "grad_norm": 0.4092175364494324,
      "learning_rate": 4.293785310734464e-06,
      "loss": 0.6578,
      "step": 760
    },
    {
      "epoch": 0.17643067179371183,
      "grad_norm": 1.4677420854568481,
      "learning_rate": 4.40677966101695e-06,
      "loss": 0.6246,
      "step": 780
    },
    {
      "epoch": 0.18095453517303778,
      "grad_norm": 0.56260085105896,
      "learning_rate": 4.519774011299436e-06,
      "loss": 0.6593,
      "step": 800
    },
    {
      "epoch": 0.18547839855236373,
      "grad_norm": 0.6638891696929932,
      "learning_rate": 4.632768361581922e-06,
      "loss": 0.6385,
      "step": 820
    },
    {
      "epoch": 0.19000226193168968,
      "grad_norm": 0.9223029613494873,
      "learning_rate": 4.745762711864408e-06,
      "loss": 0.6267,
      "step": 840
    },
    {
      "epoch": 0.1945261253110156,
      "grad_norm": 0.666900634765625,
      "learning_rate": 4.8587570621468936e-06,
      "loss": 0.5532,
      "step": 860
    },
    {
      "epoch": 0.19904998869034154,
      "grad_norm": 0.7605287432670593,
      "learning_rate": 4.9717514124293796e-06,
      "loss": 0.6349,
      "step": 880
    },
    {
      "epoch": 0.2035738520696675,
      "grad_norm": 0.7302986979484558,
      "learning_rate": 4.9997779957257335e-06,
      "loss": 0.6167,
      "step": 900
    },
    {
      "epoch": 0.20809771544899344,
      "grad_norm": 0.9552111625671387,
      "learning_rate": 4.998791389568115e-06,
      "loss": 0.6314,
      "step": 920
    },
    {
      "epoch": 0.2126215788283194,
      "grad_norm": 1.1077909469604492,
      "learning_rate": 4.997015825567898e-06,
      "loss": 0.6525,
      "step": 940
    },
    {
      "epoch": 0.21714544220764534,
      "grad_norm": 0.6709986925125122,
      "learning_rate": 4.994451864334162e-06,
      "loss": 0.5795,
      "step": 960
    },
    {
      "epoch": 0.22166930558697126,
      "grad_norm": 0.5601279735565186,
      "learning_rate": 4.991100315401203e-06,
      "loss": 0.6191,
      "step": 980
    },
    {
      "epoch": 0.2261931689662972,
      "grad_norm": 1.1237138509750366,
      "learning_rate": 4.986962236972929e-06,
      "loss": 0.6329,
      "step": 1000
    },
    {
      "epoch": 0.23071703234562316,
      "grad_norm": 0.987455427646637,
      "learning_rate": 4.982038935588749e-06,
      "loss": 0.572,
      "step": 1020
    },
    {
      "epoch": 0.2352408957249491,
      "grad_norm": 1.0568902492523193,
      "learning_rate": 4.976331965711056e-06,
      "loss": 0.6261,
      "step": 1040
    },
    {
      "epoch": 0.23976475910427505,
      "grad_norm": 0.5604099035263062,
      "learning_rate": 4.969843129234422e-06,
      "loss": 0.554,
      "step": 1060
    },
    {
      "epoch": 0.244288622483601,
      "grad_norm": 0.9130414128303528,
      "learning_rate": 4.962574474916678e-06,
      "loss": 0.5794,
      "step": 1080
    },
    {
      "epoch": 0.24881248586292695,
      "grad_norm": 1.5077348947525024,
      "learning_rate": 4.954528297732047e-06,
      "loss": 0.6414,
      "step": 1100
    },
    {
      "epoch": 0.2533363492422529,
      "grad_norm": 0.5713602900505066,
      "learning_rate": 4.94570713814654e-06,
      "loss": 0.6231,
      "step": 1120
    },
    {
      "epoch": 0.2578602126215788,
      "grad_norm": 1.139797568321228,
      "learning_rate": 4.93611378131584e-06,
      "loss": 0.6535,
      "step": 1140
    },
    {
      "epoch": 0.2623840760009048,
      "grad_norm": 1.476881980895996,
      "learning_rate": 4.925751256205924e-06,
      "loss": 0.561,
      "step": 1160
    },
    {
      "epoch": 0.2669079393802307,
      "grad_norm": 0.9005554914474487,
      "learning_rate": 4.914622834636717e-06,
      "loss": 0.6006,
      "step": 1180
    },
    {
      "epoch": 0.27143180275955664,
      "grad_norm": 0.6789364218711853,
      "learning_rate": 4.902732030249057e-06,
      "loss": 0.6368,
      "step": 1200
    },
    {
      "epoch": 0.2759556661388826,
      "grad_norm": 0.8574463725090027,
      "learning_rate": 4.890082597395316e-06,
      "loss": 0.6138,
      "step": 1220
    },
    {
      "epoch": 0.28047952951820854,
      "grad_norm": 1.226874589920044,
      "learning_rate": 4.876678529954011e-06,
      "loss": 0.618,
      "step": 1240
    },
    {
      "epoch": 0.2850033928975345,
      "grad_norm": 0.49481701850891113,
      "learning_rate": 4.8625240600688e-06,
      "loss": 0.5962,
      "step": 1260
    },
    {
      "epoch": 0.28952725627686043,
      "grad_norm": 0.8259846568107605,
      "learning_rate": 4.847623656812241e-06,
      "loss": 0.5725,
      "step": 1280
    },
    {
      "epoch": 0.2940511196561864,
      "grad_norm": 1.3541384935379028,
      "learning_rate": 4.831982024774742e-06,
      "loss": 0.6237,
      "step": 1300
    },
    {
      "epoch": 0.29857498303551233,
      "grad_norm": 0.6489030718803406,
      "learning_rate": 4.815604102579159e-06,
      "loss": 0.6559,
      "step": 1320
    },
    {
      "epoch": 0.30309884641483825,
      "grad_norm": 0.6666083931922913,
      "learning_rate": 4.798495061321492e-06,
      "loss": 0.5905,
      "step": 1340
    },
    {
      "epoch": 0.30762270979416423,
      "grad_norm": 1.033155918121338,
      "learning_rate": 4.780660302938188e-06,
      "loss": 0.5638,
      "step": 1360
    },
    {
      "epoch": 0.31214657317349015,
      "grad_norm": 0.6378616690635681,
      "learning_rate": 4.762105458500556e-06,
      "loss": 0.5433,
      "step": 1380
    },
    {
      "epoch": 0.3166704365528161,
      "grad_norm": 1.3384662866592407,
      "learning_rate": 4.742836386436831e-06,
      "loss": 0.5535,
      "step": 1400
    },
    {
      "epoch": 0.32119429993214205,
      "grad_norm": 0.6671473979949951,
      "learning_rate": 4.722859170682468e-06,
      "loss": 0.612,
      "step": 1420
    },
    {
      "epoch": 0.32571816331146797,
      "grad_norm": 0.98403000831604,
      "learning_rate": 4.702180118759219e-06,
      "loss": 0.5735,
      "step": 1440
    },
    {
      "epoch": 0.33024202669079394,
      "grad_norm": 1.0076704025268555,
      "learning_rate": 4.680805759783622e-06,
      "loss": 0.5879,
      "step": 1460
    },
    {
      "epoch": 0.33476589007011986,
      "grad_norm": 0.9248991012573242,
      "learning_rate": 4.6587428424055324e-06,
      "loss": 0.5747,
      "step": 1480
    },
    {
      "epoch": 0.33928975344944584,
      "grad_norm": 0.7047035098075867,
      "learning_rate": 4.635998332677325e-06,
      "loss": 0.5827,
      "step": 1500
    },
    {
      "epoch": 0.34381361682877176,
      "grad_norm": 0.9783876538276672,
      "learning_rate": 4.612579411854468e-06,
      "loss": 0.5671,
      "step": 1520
    },
    {
      "epoch": 0.34833748020809774,
      "grad_norm": 0.8747379183769226,
      "learning_rate": 4.588493474128137e-06,
      "loss": 0.5934,
      "step": 1540
    },
    {
      "epoch": 0.35286134358742366,
      "grad_norm": 0.8138028383255005,
      "learning_rate": 4.563748124290612e-06,
      "loss": 0.7145,
      "step": 1560
    },
    {
      "epoch": 0.3573852069667496,
      "grad_norm": 0.5983314514160156,
      "learning_rate": 4.538351175334167e-06,
      "loss": 0.5961,
      "step": 1580
    },
    {
      "epoch": 0.36190907034607556,
      "grad_norm": 0.7765645980834961,
      "learning_rate": 4.512310645984231e-06,
      "loss": 0.6588,
      "step": 1600
    },
    {
      "epoch": 0.3664329337254015,
      "grad_norm": 0.6621695756912231,
      "learning_rate": 4.485634758167589e-06,
      "loss": 0.6608,
      "step": 1620
    },
    {
      "epoch": 0.37095679710472745,
      "grad_norm": 0.5030168294906616,
      "learning_rate": 4.458331934416427e-06,
      "loss": 0.5647,
      "step": 1640
    },
    {
      "epoch": 0.3754806604840534,
      "grad_norm": 0.7155460715293884,
      "learning_rate": 4.430410795209031e-06,
      "loss": 0.5815,
      "step": 1660
    },
    {
      "epoch": 0.38000452386337935,
      "grad_norm": 1.0417250394821167,
      "learning_rate": 4.401880156248006e-06,
      "loss": 0.6304,
      "step": 1680
    },
    {
      "epoch": 0.38452838724270527,
      "grad_norm": 1.251464605331421,
      "learning_rate": 4.372749025676829e-06,
      "loss": 0.5852,
      "step": 1700
    },
    {
      "epoch": 0.3890522506220312,
      "grad_norm": 0.53974848985672,
      "learning_rate": 4.343026601235666e-06,
      "loss": 0.5883,
      "step": 1720
    },
    {
      "epoch": 0.39357611400135717,
      "grad_norm": 0.720723569393158,
      "learning_rate": 4.312722267357311e-06,
      "loss": 0.6516,
      "step": 1740
    },
    {
      "epoch": 0.3980999773806831,
      "grad_norm": 0.8760417103767395,
      "learning_rate": 4.281845592204188e-06,
      "loss": 0.5926,
      "step": 1760
    },
    {
      "epoch": 0.40262384076000907,
      "grad_norm": 0.7676084637641907,
      "learning_rate": 4.2504063246473395e-06,
      "loss": 0.6594,
      "step": 1780
    },
    {
      "epoch": 0.407147704139335,
      "grad_norm": 0.6235904097557068,
      "learning_rate": 4.218414391188353e-06,
      "loss": 0.6249,
      "step": 1800
    },
    {
      "epoch": 0.41167156751866096,
      "grad_norm": 0.5731521248817444,
      "learning_rate": 4.185879892825222e-06,
      "loss": 0.613,
      "step": 1820
    },
    {
      "epoch": 0.4161954308979869,
      "grad_norm": 0.8284727334976196,
      "learning_rate": 4.152813101863087e-06,
      "loss": 0.6487,
      "step": 1840
    },
    {
      "epoch": 0.4207192942773128,
      "grad_norm": 0.5617169141769409,
      "learning_rate": 4.119224458670905e-06,
      "loss": 0.5878,
      "step": 1860
    },
    {
      "epoch": 0.4252431576566388,
      "grad_norm": 0.647582471370697,
      "learning_rate": 4.085124568385057e-06,
      "loss": 0.6037,
      "step": 1880
    },
    {
      "epoch": 0.4297670210359647,
      "grad_norm": 0.7462798357009888,
      "learning_rate": 4.050524197560921e-06,
      "loss": 0.5911,
      "step": 1900
    },
    {
      "epoch": 0.4342908844152907,
      "grad_norm": 0.6552990674972534,
      "learning_rate": 4.015434270773485e-06,
      "loss": 0.6064,
      "step": 1920
    },
    {
      "epoch": 0.4388147477946166,
      "grad_norm": 0.8512873649597168,
      "learning_rate": 3.979865867168071e-06,
      "loss": 0.5766,
      "step": 1940
    },
    {
      "epoch": 0.4433386111739425,
      "grad_norm": 0.6242995262145996,
      "learning_rate": 3.943830216962254e-06,
      "loss": 0.603,
      "step": 1960
    },
    {
      "epoch": 0.4478624745532685,
      "grad_norm": 0.933954119682312,
      "learning_rate": 3.9073386979000715e-06,
      "loss": 0.6316,
      "step": 1980
    },
    {
      "epoch": 0.4523863379325944,
      "grad_norm": 0.5889476537704468,
      "learning_rate": 3.870402831659676e-06,
      "loss": 0.5686,
      "step": 2000
    },
    {
      "epoch": 0.4569102013119204,
      "grad_norm": 0.5107278227806091,
      "learning_rate": 3.833034280215518e-06,
      "loss": 0.6149,
      "step": 2020
    },
    {
      "epoch": 0.4614340646912463,
      "grad_norm": 0.546362042427063,
      "learning_rate": 3.795244842156246e-06,
      "loss": 0.6593,
      "step": 2040
    },
    {
      "epoch": 0.4659579280705723,
      "grad_norm": 0.7260658740997314,
      "learning_rate": 3.757046448959474e-06,
      "loss": 0.576,
      "step": 2060
    },
    {
      "epoch": 0.4704817914498982,
      "grad_norm": 0.5332072377204895,
      "learning_rate": 3.718451161224584e-06,
      "loss": 0.6064,
      "step": 2080
    },
    {
      "epoch": 0.47500565482922413,
      "grad_norm": 0.5485090613365173,
      "learning_rate": 3.679471164864763e-06,
      "loss": 0.6147,
      "step": 2100
    },
    {
      "epoch": 0.4795295182085501,
      "grad_norm": 0.6777561902999878,
      "learning_rate": 3.640118767259474e-06,
      "loss": 0.5975,
      "step": 2120
    },
    {
      "epoch": 0.48405338158787603,
      "grad_norm": 0.5616791248321533,
      "learning_rate": 3.600406393368577e-06,
      "loss": 0.5684,
      "step": 2140
    },
    {
      "epoch": 0.488577244967202,
      "grad_norm": 0.7273005843162537,
      "learning_rate": 3.560346581809328e-06,
      "loss": 0.5439,
      "step": 2160
    },
    {
      "epoch": 0.4931011083465279,
      "grad_norm": 0.399677574634552,
      "learning_rate": 3.51995198089748e-06,
      "loss": 0.6252,
      "step": 2180
    },
    {
      "epoch": 0.4976249717258539,
      "grad_norm": 0.5627871155738831,
      "learning_rate": 3.479235344653767e-06,
      "loss": 0.6454,
      "step": 2200
    },
    {
      "epoch": 0.5021488351051798,
      "grad_norm": 0.7673916816711426,
      "learning_rate": 3.4382095287769962e-06,
      "loss": 0.5485,
      "step": 2220
    },
    {
      "epoch": 0.5066726984845058,
      "grad_norm": 0.8236356377601624,
      "learning_rate": 3.396887486585047e-06,
      "loss": 0.6054,
      "step": 2240
    },
    {
      "epoch": 0.5111965618638317,
      "grad_norm": 0.9059731364250183,
      "learning_rate": 3.355282264925047e-06,
      "loss": 0.5705,
      "step": 2260
    },
    {
      "epoch": 0.5157204252431576,
      "grad_norm": 0.8679450750350952,
      "learning_rate": 3.3134070000540114e-06,
      "loss": 0.5901,
      "step": 2280
    },
    {
      "epoch": 0.5202442886224836,
      "grad_norm": 0.5972412824630737,
      "learning_rate": 3.2712749134912582e-06,
      "loss": 0.5561,
      "step": 2300
    },
    {
      "epoch": 0.5247681520018096,
      "grad_norm": 0.830877959728241,
      "learning_rate": 3.2288993078439034e-06,
      "loss": 0.6464,
      "step": 2320
    },
    {
      "epoch": 0.5292920153811355,
      "grad_norm": 1.2185938358306885,
      "learning_rate": 3.1862935626067492e-06,
      "loss": 0.5784,
      "step": 2340
    },
    {
      "epoch": 0.5338158787604614,
      "grad_norm": 0.6374582648277283,
      "learning_rate": 3.1434711299379025e-06,
      "loss": 0.5661,
      "step": 2360
    },
    {
      "epoch": 0.5383397421397874,
      "grad_norm": 0.7829556465148926,
      "learning_rate": 3.100445530411442e-06,
      "loss": 0.5385,
      "step": 2380
    },
    {
      "epoch": 0.5428636055191133,
      "grad_norm": 0.8443902134895325,
      "learning_rate": 3.0572303487484935e-06,
      "loss": 0.5565,
      "step": 2400
    },
    {
      "epoch": 0.5473874688984393,
      "grad_norm": 1.1685762405395508,
      "learning_rate": 3.013839229528046e-06,
      "loss": 0.5781,
      "step": 2420
    },
    {
      "epoch": 0.5519113322777652,
      "grad_norm": 0.9711598753929138,
      "learning_rate": 2.9702858728788714e-06,
      "loss": 0.6448,
      "step": 2440
    },
    {
      "epoch": 0.5564351956570912,
      "grad_norm": 0.43809574842453003,
      "learning_rate": 2.9265840301538995e-06,
      "loss": 0.558,
      "step": 2460
    },
    {
      "epoch": 0.5609590590364171,
      "grad_norm": 0.6405979990959167,
      "learning_rate": 2.8827474995884304e-06,
      "loss": 0.6875,
      "step": 2480
    },
    {
      "epoch": 0.565482922415743,
      "grad_norm": 0.6583331823348999,
      "learning_rate": 2.8387901219435298e-06,
      "loss": 0.5485,
      "step": 2500
    },
    {
      "epoch": 0.570006785795069,
      "grad_norm": 0.7614380121231079,
      "learning_rate": 2.7947257761360136e-06,
      "loss": 0.5632,
      "step": 2520
    },
    {
      "epoch": 0.5745306491743949,
      "grad_norm": 0.69844651222229,
      "learning_rate": 2.750568374856372e-06,
      "loss": 0.6059,
      "step": 2540
    },
    {
      "epoch": 0.5790545125537209,
      "grad_norm": 0.7071551084518433,
      "learning_rate": 2.7063318601760386e-06,
      "loss": 0.5265,
      "step": 2560
    },
    {
      "epoch": 0.5835783759330468,
      "grad_norm": 0.6122928857803345,
      "learning_rate": 2.6620301991453755e-06,
      "loss": 0.5629,
      "step": 2580
    },
    {
      "epoch": 0.5881022393123728,
      "grad_norm": 0.8555611968040466,
      "learning_rate": 2.617677379383781e-06,
      "loss": 0.5967,
      "step": 2600
    },
    {
      "epoch": 0.5926261026916987,
      "grad_norm": 0.5617034435272217,
      "learning_rate": 2.5732874046632938e-06,
      "loss": 0.6009,
      "step": 2620
    },
    {
      "epoch": 0.5971499660710247,
      "grad_norm": 0.6981135606765747,
      "learning_rate": 2.528874290487102e-06,
      "loss": 0.5641,
      "step": 2640
    },
    {
      "epoch": 0.6016738294503506,
      "grad_norm": 0.5575506687164307,
      "learning_rate": 2.4844520596643495e-06,
      "loss": 0.5759,
      "step": 2660
    },
    {
      "epoch": 0.6061976928296765,
      "grad_norm": 0.7685838341712952,
      "learning_rate": 2.4400347378826296e-06,
      "loss": 0.5343,
      "step": 2680
    },
    {
      "epoch": 0.6107215562090025,
      "grad_norm": 0.7420554161071777,
      "learning_rate": 2.395636349279577e-06,
      "loss": 0.6218,
      "step": 2700
    },
    {
      "epoch": 0.6152454195883285,
      "grad_norm": 0.43521085381507874,
      "learning_rate": 2.3512709120149436e-06,
      "loss": 0.5303,
      "step": 2720
    },
    {
      "epoch": 0.6197692829676543,
      "grad_norm": 0.6557121276855469,
      "learning_rate": 2.3069524338445666e-06,
      "loss": 0.5794,
      "step": 2740
    },
    {
      "epoch": 0.6242931463469803,
      "grad_norm": 0.7263210415840149,
      "learning_rate": 2.2626949076976153e-06,
      "loss": 0.6239,
      "step": 2760
    },
    {
      "epoch": 0.6288170097263063,
      "grad_norm": 0.5805336833000183,
      "learning_rate": 2.2185123072585253e-06,
      "loss": 0.5821,
      "step": 2780
    },
    {
      "epoch": 0.6333408731056323,
      "grad_norm": 0.8271047472953796,
      "learning_rate": 2.174418582555005e-06,
      "loss": 0.5924,
      "step": 2800
    },
    {
      "epoch": 0.6378647364849581,
      "grad_norm": 0.49003851413726807,
      "learning_rate": 2.1304276555535123e-06,
      "loss": 0.5567,
      "step": 2820
    },
    {
      "epoch": 0.6423885998642841,
      "grad_norm": 0.6948228478431702,
      "learning_rate": 2.086553415763597e-06,
      "loss": 0.541,
      "step": 2840
    },
    {
      "epoch": 0.6469124632436101,
      "grad_norm": 0.7717708945274353,
      "learning_rate": 2.0428097158524768e-06,
      "loss": 0.6128,
      "step": 2860
    },
    {
      "epoch": 0.6514363266229359,
      "grad_norm": 0.8652535676956177,
      "learning_rate": 1.999210367271261e-06,
      "loss": 0.5484,
      "step": 2880
    },
    {
      "epoch": 0.6559601900022619,
      "grad_norm": 0.48044684529304504,
      "learning_rate": 1.9557691358941796e-06,
      "loss": 0.6335,
      "step": 2900
    },
    {
      "epoch": 0.6604840533815879,
      "grad_norm": 0.6042224764823914,
      "learning_rate": 1.9124997376722038e-06,
      "loss": 0.5719,
      "step": 2920
    },
    {
      "epoch": 0.6650079167609139,
      "grad_norm": 0.45004037022590637,
      "learning_rate": 1.869415834302429e-06,
      "loss": 0.5828,
      "step": 2940
    },
    {
      "epoch": 0.6695317801402397,
      "grad_norm": 0.7315308451652527,
      "learning_rate": 1.826531028914593e-06,
      "loss": 0.5809,
      "step": 2960
    },
    {
      "epoch": 0.6740556435195657,
      "grad_norm": 0.4399823546409607,
      "learning_rate": 1.7838588617760733e-06,
      "loss": 0.5933,
      "step": 2980
    },
    {
      "epoch": 0.6785795068988917,
      "grad_norm": 1.064976692199707,
      "learning_rate": 1.7414128060167457e-06,
      "loss": 0.5469,
      "step": 3000
    },
    {
      "epoch": 0.6831033702782175,
      "grad_norm": 0.7710141539573669,
      "learning_rate": 1.6992062633750346e-06,
      "loss": 0.585,
      "step": 3020
    },
    {
      "epoch": 0.6876272336575435,
      "grad_norm": 0.673542857170105,
      "learning_rate": 1.657252559966502e-06,
      "loss": 0.5436,
      "step": 3040
    },
    {
      "epoch": 0.6921510970368695,
      "grad_norm": 0.6577861309051514,
      "learning_rate": 1.6155649420763214e-06,
      "loss": 0.6201,
      "step": 3060
    },
    {
      "epoch": 0.6966749604161955,
      "grad_norm": 0.9896159768104553,
      "learning_rate": 1.5741565719769447e-06,
      "loss": 0.6079,
      "step": 3080
    },
    {
      "epoch": 0.7011988237955213,
      "grad_norm": 0.7078260183334351,
      "learning_rate": 1.5330405237723078e-06,
      "loss": 0.6218,
      "step": 3100
    },
    {
      "epoch": 0.7057226871748473,
      "grad_norm": 1.2103734016418457,
      "learning_rate": 1.4922297792698645e-06,
      "loss": 0.5955,
      "step": 3120
    },
    {
      "epoch": 0.7102465505541733,
      "grad_norm": 0.6530666947364807,
      "learning_rate": 1.4517372238817668e-06,
      "loss": 0.5647,
      "step": 3140
    },
    {
      "epoch": 0.7147704139334992,
      "grad_norm": 0.6449224352836609,
      "learning_rate": 1.4115756425564798e-06,
      "loss": 0.6194,
      "step": 3160
    },
    {
      "epoch": 0.7192942773128251,
      "grad_norm": 0.5960292220115662,
      "learning_rate": 1.3717577157421169e-06,
      "loss": 0.5926,
      "step": 3180
    },
    {
      "epoch": 0.7238181406921511,
      "grad_norm": 0.6450957655906677,
      "learning_rate": 1.3322960153827632e-06,
      "loss": 0.6137,
      "step": 3200
    },
    {
      "epoch": 0.7283420040714771,
      "grad_norm": 0.5211399793624878,
      "learning_rate": 1.293203000949072e-06,
      "loss": 0.5796,
      "step": 3220
    },
    {
      "epoch": 0.732865867450803,
      "grad_norm": 0.6112457513809204,
      "learning_rate": 1.2544910155043545e-06,
      "loss": 0.6187,
      "step": 3240
    },
    {
      "epoch": 0.7373897308301289,
      "grad_norm": 0.5904879570007324,
      "learning_rate": 1.2161722818074412e-06,
      "loss": 0.5708,
      "step": 3260
    },
    {
      "epoch": 0.7419135942094549,
      "grad_norm": 0.5986872911453247,
      "learning_rate": 1.1782588984535129e-06,
      "loss": 0.5455,
      "step": 3280
    },
    {
      "epoch": 0.7464374575887808,
      "grad_norm": 0.716018795967102,
      "learning_rate": 1.14076283605415e-06,
      "loss": 0.6109,
      "step": 3300
    },
    {
      "epoch": 0.7509613209681067,
      "grad_norm": 0.5863745212554932,
      "learning_rate": 1.1036959334577708e-06,
      "loss": 0.5875,
      "step": 3320
    },
    {
      "epoch": 0.7554851843474327,
      "grad_norm": 0.8205773830413818,
      "learning_rate": 1.0670698940116937e-06,
      "loss": 0.6015,
      "step": 3340
    },
    {
      "epoch": 0.7600090477267587,
      "grad_norm": 0.6910251975059509,
      "learning_rate": 1.030896281866964e-06,
      "loss": 0.5625,
      "step": 3360
    },
    {
      "epoch": 0.7645329111060846,
      "grad_norm": 0.42625927925109863,
      "learning_rate": 9.951865183271443e-07,
      "loss": 0.5823,
      "step": 3380
    },
    {
      "epoch": 0.7690567744854105,
      "grad_norm": 0.742891788482666,
      "learning_rate": 9.599518782421954e-07,
      "loss": 0.5209,
      "step": 3400
    },
    {
      "epoch": 0.7735806378647365,
      "grad_norm": 0.6565542817115784,
      "learning_rate": 9.252034864486062e-07,
      "loss": 0.5896,
      "step": 3420
    },
    {
      "epoch": 0.7781045012440624,
      "grad_norm": 0.7679634690284729,
      "learning_rate": 8.909523142568807e-07,
      "loss": 0.6889,
      "step": 3440
    },
    {
      "epoch": 0.7826283646233884,
      "grad_norm": 0.8022564649581909,
      "learning_rate": 8.572091759875081e-07,
      "loss": 0.5643,
      "step": 3460
    },
    {
      "epoch": 0.7871522280027143,
      "grad_norm": 1.0336594581604004,
      "learning_rate": 8.239847255564889e-07,
      "loss": 0.6212,
      "step": 3480
    },
    {
      "epoch": 0.7916760913820403,
      "grad_norm": 0.6547170877456665,
      "learning_rate": 7.912894531115184e-07,
      "loss": 0.5685,
      "step": 3500
    },
    {
      "epoch": 0.7961999547613662,
      "grad_norm": 0.725557267665863,
      "learning_rate": 7.591336817198683e-07,
      "loss": 0.5953,
      "step": 3520
    },
    {
      "epoch": 0.8007238181406922,
      "grad_norm": 0.8147058486938477,
      "learning_rate": 7.275275641090238e-07,
      "loss": 0.628,
      "step": 3540
    },
    {
      "epoch": 0.8052476815200181,
      "grad_norm": 0.5000850558280945,
      "learning_rate": 6.964810794611093e-07,
      "loss": 0.5696,
      "step": 3560
    },
    {
      "epoch": 0.809771544899344,
      "grad_norm": 0.9064733386039734,
      "learning_rate": 6.660040302620974e-07,
      "loss": 0.6752,
      "step": 3580
    },
    {
      "epoch": 0.81429540827867,
      "grad_norm": 0.7876561284065247,
      "learning_rate": 6.361060392068177e-07,
      "loss": 0.6045,
      "step": 3600
    },
    {
      "epoch": 0.818819271657996,
      "grad_norm": 0.5467254519462585,
      "learning_rate": 6.067965461607275e-07,
      "loss": 0.5963,
      "step": 3620
    },
    {
      "epoch": 0.8233431350373219,
      "grad_norm": 0.5970877408981323,
      "learning_rate": 5.780848051794066e-07,
      "loss": 0.6115,
      "step": 3640
    },
    {
      "epoch": 0.8278669984166478,
      "grad_norm": 0.809258759021759,
      "learning_rate": 5.499798815867208e-07,
      "loss": 0.53,
      "step": 3660
    },
    {
      "epoch": 0.8323908617959738,
      "grad_norm": 0.9609009027481079,
      "learning_rate": 5.224906491125778e-07,
      "loss": 0.5172,
      "step": 3680
    },
    {
      "epoch": 0.8369147251752997,
      "grad_norm": 0.5807494521141052,
      "learning_rate": 4.95625787091169e-07,
      "loss": 0.5784,
      "step": 3700
    },
    {
      "epoch": 0.8414385885546256,
      "grad_norm": 0.8828917145729065,
      "learning_rate": 4.693937777205973e-07,
      "loss": 0.6114,
      "step": 3720
    },
    {
      "epoch": 0.8459624519339516,
      "grad_norm": 1.2154064178466797,
      "learning_rate": 4.438029033847402e-07,
      "loss": 0.5837,
      "step": 3740
    },
    {
      "epoch": 0.8504863153132776,
      "grad_norm": 0.5537241101264954,
      "learning_rate": 4.1886124403820864e-07,
      "loss": 0.5766,
      "step": 3760
    },
    {
      "epoch": 0.8550101786926034,
      "grad_norm": 0.9004805088043213,
      "learning_rate": 3.94576674655211e-07,
      "loss": 0.6033,
      "step": 3780
    },
    {
      "epoch": 0.8595340420719294,
      "grad_norm": 0.7978100776672363,
      "learning_rate": 3.7095686274314757e-07,
      "loss": 0.5284,
      "step": 3800
    },
    {
      "epoch": 0.8640579054512554,
      "grad_norm": 0.7304428219795227,
      "learning_rate": 3.4800926592169995e-07,
      "loss": 0.6011,
      "step": 3820
    },
    {
      "epoch": 0.8685817688305814,
      "grad_norm": 0.7252260446548462,
      "learning_rate": 3.2574112956819613e-07,
      "loss": 0.5691,
      "step": 3840
    },
    {
      "epoch": 0.8731056322099072,
      "grad_norm": 0.6625013947486877,
      "learning_rate": 3.041594845299856e-07,
      "loss": 0.5782,
      "step": 3860
    },
    {
      "epoch": 0.8776294955892332,
      "grad_norm": 0.5668479800224304,
      "learning_rate": 2.832711449045461e-07,
      "loss": 0.5781,
      "step": 3880
    },
    {
      "epoch": 0.8821533589685592,
      "grad_norm": 0.6604365110397339,
      "learning_rate": 2.630827058880345e-07,
      "loss": 0.5579,
      "step": 3900
    },
    {
      "epoch": 0.886677222347885,
      "grad_norm": 0.9126925468444824,
      "learning_rate": 2.4360054169294377e-07,
      "loss": 0.6684,
      "step": 3920
    },
    {
      "epoch": 0.891201085727211,
      "grad_norm": 0.5072206854820251,
      "learning_rate": 2.2483080353553537e-07,
      "loss": 0.6443,
      "step": 3940
    },
    {
      "epoch": 0.895724949106537,
      "grad_norm": 0.5561603307723999,
      "learning_rate": 2.067794176936838e-07,
      "loss": 0.6012,
      "step": 3960
    },
    {
      "epoch": 0.900248812485863,
      "grad_norm": 0.5377748012542725,
      "learning_rate": 1.8945208363573553e-07,
      "loss": 0.5513,
      "step": 3980
    },
    {
      "epoch": 0.9047726758651888,
      "grad_norm": 0.6415522694587708,
      "learning_rate": 1.7285427222098388e-07,
      "loss": 0.6421,
      "step": 4000
    },
    {
      "epoch": 0.9092965392445148,
      "grad_norm": 0.7068675756454468,
      "learning_rate": 1.569912239723237e-07,
      "loss": 0.6059,
      "step": 4020
    },
    {
      "epoch": 0.9138204026238408,
      "grad_norm": 0.5292190909385681,
      "learning_rate": 1.4186794742162918e-07,
      "loss": 0.5233,
      "step": 4040
    },
    {
      "epoch": 0.9183442660031667,
      "grad_norm": 0.7200678586959839,
      "learning_rate": 1.27489217528384e-07,
      "loss": 0.607,
      "step": 4060
    },
    {
      "epoch": 0.9228681293824926,
      "grad_norm": 0.5470794439315796,
      "learning_rate": 1.138595741720519e-07,
      "loss": 0.6102,
      "step": 4080
    },
    {
      "epoch": 0.9273919927618186,
      "grad_norm": 0.6571109294891357,
      "learning_rate": 1.009833207186775e-07,
      "loss": 0.5965,
      "step": 4100
    },
    {
      "epoch": 0.9319158561411446,
      "grad_norm": 0.534355640411377,
      "learning_rate": 8.886452266215817e-08,
      "loss": 0.6624,
      "step": 4120
    },
    {
      "epoch": 0.9364397195204704,
      "grad_norm": 0.7418869137763977,
      "learning_rate": 7.750700634062303e-08,
      "loss": 0.6008,
      "step": 4140
    },
    {
      "epoch": 0.9409635828997964,
      "grad_norm": 0.6004541516304016,
      "learning_rate": 6.691435772831933e-08,
      "loss": 0.582,
      "step": 4160
    },
    {
      "epoch": 0.9454874462791224,
      "grad_norm": 0.48619237542152405,
      "learning_rate": 5.708992130339458e-08,
      "loss": 0.5868,
      "step": 4180
    },
    {
      "epoch": 0.9500113096584483,
      "grad_norm": 0.5165609121322632,
      "learning_rate": 4.8036798991923925e-08,
      "loss": 0.5558,
      "step": 4200
    },
    {
      "epoch": 0.9545351730377742,
      "grad_norm": 0.7379537224769592,
      "learning_rate": 3.9757849188520046e-08,
      "loss": 0.617,
      "step": 4220
    },
    {
      "epoch": 0.9590590364171002,
      "grad_norm": 0.6974282264709473,
      "learning_rate": 3.2255685853837815e-08,
      "loss": 0.5584,
      "step": 4240
    },
    {
      "epoch": 0.9635828997964262,
      "grad_norm": 0.6777763366699219,
      "learning_rate": 2.553267768925116e-08,
      "loss": 0.553,
      "step": 4260
    },
    {
      "epoch": 0.9681067631757521,
      "grad_norm": 1.0349258184432983,
      "learning_rate": 1.959094738897216e-08,
      "loss": 0.5674,
      "step": 4280
    },
    {
      "epoch": 0.972630626555078,
      "grad_norm": 1.0187406539916992,
      "learning_rate": 1.4432370969836851e-08,
      "loss": 0.6268,
      "step": 4300
    },
    {
      "epoch": 0.977154489934404,
      "grad_norm": 0.4671124219894409,
      "learning_rate": 1.0058577178982954e-08,
      "loss": 0.6393,
      "step": 4320
    },
    {
      "epoch": 0.9816783533137299,
      "grad_norm": 0.6963977217674255,
      "learning_rate": 6.4709469795945505e-09,
      "loss": 0.6028,
      "step": 4340
    },
    {
      "epoch": 0.9862022166930559,
      "grad_norm": 0.7137852907180786,
      "learning_rate": 3.67061311488226e-09,
      "loss": 0.5654,
      "step": 4360
    },
    {
      "epoch": 0.9907260800723818,
      "grad_norm": 0.6567050218582153,
      "learning_rate": 1.6584597504365475e-09,
      "loss": 0.5807,
      "step": 4380
    },
    {
      "epoch": 0.9952499434517078,
      "grad_norm": 0.9630624651908875,
      "learning_rate": 4.3512219506519626e-10,
      "loss": 0.5787,
      "step": 4400
    }
  ],
  "logging_steps": 20,
  "max_steps": 4421,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1225101215714509e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}