{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9981634527089072,
  "eval_steps": 500,
  "global_step": 1088,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.018365472910927456,
      "grad_norm": 0.27851340305035927,
      "learning_rate": 9.997915740506687e-05,
      "loss": 1.0445,
      "step": 10
    },
    {
      "epoch": 0.03673094582185491,
      "grad_norm": 0.061819145479423056,
      "learning_rate": 9.991664699681799e-05,
      "loss": 0.5539,
      "step": 20
    },
    {
      "epoch": 0.05509641873278237,
      "grad_norm": 0.03760363481720962,
      "learning_rate": 9.981252089041809e-05,
      "loss": 0.4787,
      "step": 30
    },
    {
      "epoch": 0.07346189164370982,
      "grad_norm": 0.0313950706656064,
      "learning_rate": 9.966686589619751e-05,
      "loss": 0.468,
      "step": 40
    },
    {
      "epoch": 0.09182736455463728,
      "grad_norm": 0.022877630980033918,
      "learning_rate": 9.947980344727799e-05,
      "loss": 0.4451,
      "step": 50
    },
    {
      "epoch": 0.11019283746556474,
      "grad_norm": 0.029933452784706518,
      "learning_rate": 9.925148949833355e-05,
      "loss": 0.4469,
      "step": 60
    },
    {
      "epoch": 0.1285583103764922,
      "grad_norm": 0.023764995733481963,
      "learning_rate": 9.898211439557042e-05,
      "loss": 0.4424,
      "step": 70
    },
    {
      "epoch": 0.14692378328741965,
      "grad_norm": 0.02327001477570512,
      "learning_rate": 9.867190271803465e-05,
      "loss": 0.4336,
      "step": 80
    },
    {
      "epoch": 0.1652892561983471,
      "grad_norm": 0.024539265266611217,
      "learning_rate": 9.832111309037979e-05,
      "loss": 0.4321,
      "step": 90
    },
    {
      "epoch": 0.18365472910927455,
      "grad_norm": 0.039657236107179715,
      "learning_rate": 9.793003796725048e-05,
      "loss": 0.4384,
      "step": 100
    },
    {
      "epoch": 0.20202020202020202,
      "grad_norm": 0.027774934175057744,
      "learning_rate": 9.749900338946192e-05,
      "loss": 0.4283,
      "step": 110
    },
    {
      "epoch": 0.22038567493112948,
      "grad_norm": 0.02480612425885802,
      "learning_rate": 9.702836871217839e-05,
      "loss": 0.4327,
      "step": 120
    },
    {
      "epoch": 0.23875114784205692,
      "grad_norm": 0.025217057933479833,
      "learning_rate": 9.651852630531748e-05,
      "loss": 0.4191,
      "step": 130
    },
    {
      "epoch": 0.2571166207529844,
      "grad_norm": 0.042028378260899814,
      "learning_rate": 9.596990122642983e-05,
      "loss": 0.4171,
      "step": 140
    },
    {
      "epoch": 0.27548209366391185,
      "grad_norm": 0.030731629233234352,
      "learning_rate": 9.538295086632703e-05,
      "loss": 0.4273,
      "step": 150
    },
    {
      "epoch": 0.2938475665748393,
      "grad_norm": 0.06006724597666194,
      "learning_rate": 9.475816456775313e-05,
      "loss": 0.4266,
      "step": 160
    },
    {
      "epoch": 0.3122130394857668,
      "grad_norm": 0.02893865752550571,
      "learning_rate": 9.409606321741775e-05,
      "loss": 0.4196,
      "step": 170
    },
    {
      "epoch": 0.3305785123966942,
      "grad_norm": 0.026458687435744248,
      "learning_rate": 9.339719881173093e-05,
      "loss": 0.423,
      "step": 180
    },
    {
      "epoch": 0.34894398530762166,
      "grad_norm": 0.03576661132708211,
      "learning_rate": 9.266215399660146e-05,
      "loss": 0.4192,
      "step": 190
    },
    {
      "epoch": 0.3673094582185491,
      "grad_norm": 0.02652249141257209,
      "learning_rate": 9.189154158168292e-05,
      "loss": 0.4164,
      "step": 200
    },
    {
      "epoch": 0.3856749311294766,
      "grad_norm": 0.025216448544065753,
      "learning_rate": 9.108600402947192e-05,
      "loss": 0.4272,
      "step": 210
    },
    {
      "epoch": 0.40404040404040403,
      "grad_norm": 0.026533959286346408,
      "learning_rate": 9.02462129196846e-05,
      "loss": 0.4189,
      "step": 220
    },
    {
      "epoch": 0.42240587695133147,
      "grad_norm": 0.026029222217085504,
      "learning_rate": 8.93728683893582e-05,
      "loss": 0.4207,
      "step": 230
    },
    {
      "epoch": 0.44077134986225897,
      "grad_norm": 0.028412018887179133,
      "learning_rate": 8.846669854914396e-05,
      "loss": 0.417,
      "step": 240
    },
    {
      "epoch": 0.4591368227731864,
      "grad_norm": 0.026783824853837783,
      "learning_rate": 8.752845887627872e-05,
      "loss": 0.4147,
      "step": 250
    },
    {
      "epoch": 0.47750229568411384,
      "grad_norm": 0.026947377894842927,
      "learning_rate": 8.655893158474055e-05,
      "loss": 0.4179,
      "step": 260
    },
    {
      "epoch": 0.49586776859504134,
      "grad_norm": 0.03613738404815021,
      "learning_rate": 8.555892497311402e-05,
      "loss": 0.4068,
      "step": 270
    },
    {
      "epoch": 0.5142332415059688,
      "grad_norm": 0.027745465710559532,
      "learning_rate": 8.452927275070858e-05,
      "loss": 0.4103,
      "step": 280
    },
    {
      "epoch": 0.5325987144168962,
      "grad_norm": 0.025444027330206734,
      "learning_rate": 8.347083334249199e-05,
      "loss": 0.4009,
      "step": 290
    },
    {
      "epoch": 0.5509641873278237,
      "grad_norm": 0.029718700126100683,
      "learning_rate": 8.23844891734181e-05,
      "loss": 0.4183,
      "step": 300
    },
    {
      "epoch": 0.5693296602387512,
      "grad_norm": 0.029941432231859456,
      "learning_rate": 8.12711459327459e-05,
      "loss": 0.4199,
      "step": 310
    },
    {
      "epoch": 0.5876951331496786,
      "grad_norm": 0.02728726308576057,
      "learning_rate": 8.013173181896283e-05,
      "loss": 0.4023,
      "step": 320
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.02842461149504098,
      "learning_rate": 7.89671967659423e-05,
      "loss": 0.4185,
      "step": 330
    },
    {
      "epoch": 0.6244260789715336,
      "grad_norm": 0.02557752553781187,
      "learning_rate": 7.777851165098012e-05,
      "loss": 0.4259,
      "step": 340
    },
    {
      "epoch": 0.642791551882461,
      "grad_norm": 0.026972728188092137,
      "learning_rate": 7.656666748537045e-05,
      "loss": 0.4039,
      "step": 350
    },
    {
      "epoch": 0.6611570247933884,
      "grad_norm": 0.024371307606132524,
      "learning_rate": 7.533267458819598e-05,
      "loss": 0.4216,
      "step": 360
    },
    {
      "epoch": 0.6795224977043158,
      "grad_norm": 0.031096864607523835,
      "learning_rate": 7.407756174402089e-05,
      "loss": 0.4112,
      "step": 370
    },
    {
      "epoch": 0.6978879706152433,
      "grad_norm": 0.027930614232647415,
      "learning_rate": 7.280237534518947e-05,
      "loss": 0.4154,
      "step": 380
    },
    {
      "epoch": 0.7162534435261708,
      "grad_norm": 0.02788401233462263,
      "learning_rate": 7.150817851944472e-05,
      "loss": 0.4079,
      "step": 390
    },
    {
      "epoch": 0.7346189164370982,
      "grad_norm": 0.030760734784571572,
      "learning_rate": 7.019605024359474e-05,
      "loss": 0.411,
      "step": 400
    },
    {
      "epoch": 0.7529843893480257,
      "grad_norm": 0.025453805261318345,
      "learning_rate": 6.886708444396573e-05,
      "loss": 0.4065,
      "step": 410
    },
    {
      "epoch": 0.7713498622589532,
      "grad_norm": 0.028933030043804563,
      "learning_rate": 6.75223890843913e-05,
      "loss": 0.4171,
      "step": 420
    },
    {
      "epoch": 0.7897153351698806,
      "grad_norm": 0.026727439486227408,
      "learning_rate": 6.6163085242499e-05,
      "loss": 0.408,
      "step": 430
    },
    {
      "epoch": 0.8080808080808081,
      "grad_norm": 0.028829235828263126,
      "learning_rate": 6.479030617506353e-05,
      "loss": 0.4148,
      "step": 440
    },
    {
      "epoch": 0.8264462809917356,
      "grad_norm": 0.02837397571927983,
      "learning_rate": 6.34051963732063e-05,
      "loss": 0.4142,
      "step": 450
    },
    {
      "epoch": 0.8448117539026629,
      "grad_norm": 0.030025509075918014,
      "learning_rate": 6.200891060822883e-05,
      "loss": 0.4248,
      "step": 460
    },
    {
      "epoch": 0.8631772268135904,
      "grad_norm": 0.02918663842326674,
      "learning_rate": 6.060261296887554e-05,
      "loss": 0.4106,
      "step": 470
    },
    {
      "epoch": 0.8815426997245179,
      "grad_norm": 0.030657261653125826,
      "learning_rate": 5.918747589082853e-05,
      "loss": 0.4118,
      "step": 480
    },
    {
      "epoch": 0.8999081726354453,
      "grad_norm": 0.029002395463699617,
      "learning_rate": 5.776467917924348e-05,
      "loss": 0.409,
      "step": 490
    },
    {
      "epoch": 0.9182736455463728,
      "grad_norm": 0.02963746695292411,
      "learning_rate": 5.6335409025141694e-05,
      "loss": 0.406,
      "step": 500
    },
    {
      "epoch": 0.9366391184573003,
      "grad_norm": 0.03089177834981146,
      "learning_rate": 5.490085701647805e-05,
      "loss": 0.4102,
      "step": 510
    },
    {
      "epoch": 0.9550045913682277,
      "grad_norm": 0.032695907122492804,
      "learning_rate": 5.346221914470959e-05,
      "loss": 0.4025,
      "step": 520
    },
    {
      "epoch": 0.9733700642791552,
      "grad_norm": 0.0283784844379621,
      "learning_rate": 5.2020694807693015e-05,
      "loss": 0.401,
      "step": 530
    },
    {
      "epoch": 0.9917355371900827,
      "grad_norm": 0.02859989539113277,
      "learning_rate": 5.0577485809742044e-05,
      "loss": 0.4163,
      "step": 540
    },
    {
      "epoch": 1.0101010101010102,
      "grad_norm": 0.03103627192686722,
      "learning_rate": 4.913379535967859e-05,
      "loss": 0.4039,
      "step": 550
    },
    {
      "epoch": 1.0284664830119377,
      "grad_norm": 0.03322012900664279,
      "learning_rate": 4.7690827067713035e-05,
      "loss": 0.3914,
      "step": 560
    },
    {
      "epoch": 1.046831955922865,
      "grad_norm": 0.03642666862930741,
      "learning_rate": 4.6249783941989785e-05,
      "loss": 0.3908,
      "step": 570
    },
    {
      "epoch": 1.0651974288337924,
      "grad_norm": 0.0336389842141535,
      "learning_rate": 4.481186738563492e-05,
      "loss": 0.3916,
      "step": 580
    },
    {
      "epoch": 1.08356290174472,
      "grad_norm": 0.03449706815272795,
      "learning_rate": 4.3378276195141665e-05,
      "loss": 0.3858,
      "step": 590
    },
    {
      "epoch": 1.1019283746556474,
      "grad_norm": 0.036410754456203426,
      "learning_rate": 4.195020556092935e-05,
      "loss": 0.3837,
      "step": 600
    },
    {
      "epoch": 1.120293847566575,
      "grad_norm": 0.03485127798044261,
      "learning_rate": 4.05288460709086e-05,
      "loss": 0.3904,
      "step": 610
    },
    {
      "epoch": 1.1386593204775024,
      "grad_norm": 0.038047760695129655,
      "learning_rate": 3.911538271788358e-05,
      "loss": 0.3964,
      "step": 620
    },
    {
      "epoch": 1.1570247933884297,
      "grad_norm": 0.038412283865629844,
      "learning_rate": 3.7710993911619094e-05,
      "loss": 0.3774,
      "step": 630
    },
    {
      "epoch": 1.1753902662993572,
      "grad_norm": 0.036371859571502144,
      "learning_rate": 3.631685049639586e-05,
      "loss": 0.3907,
      "step": 640
    },
    {
      "epoch": 1.1937557392102847,
      "grad_norm": 0.03976682097907334,
      "learning_rate": 3.493411477487315e-05,
      "loss": 0.3858,
      "step": 650
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.037005097607228386,
      "learning_rate": 3.3563939539072705e-05,
      "loss": 0.3847,
      "step": 660
    },
    {
      "epoch": 1.2304866850321396,
      "grad_norm": 0.03996197909827653,
      "learning_rate": 3.2207467109291586e-05,
      "loss": 0.3979,
      "step": 670
    },
    {
      "epoch": 1.248852157943067,
      "grad_norm": 0.04158286829796218,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.39,
      "step": 680
    },
    {
      "epoch": 1.2672176308539944,
      "grad_norm": 0.04139047858760264,
      "learning_rate": 2.9540141885736262e-05,
      "loss": 0.3874,
      "step": 690
    },
    {
      "epoch": 1.285583103764922,
      "grad_norm": 0.041594762089601535,
      "learning_rate": 2.8231512851129593e-05,
      "loss": 0.3934,
      "step": 700
    },
    {
      "epoch": 1.3039485766758494,
      "grad_norm": 0.0431336076994283,
      "learning_rate": 2.6941032286920985e-05,
      "loss": 0.3775,
      "step": 710
    },
    {
      "epoch": 1.322314049586777,
      "grad_norm": 0.04035918987424071,
      "learning_rate": 2.5669776071657192e-05,
      "loss": 0.4017,
      "step": 720
    },
    {
      "epoch": 1.3406795224977044,
      "grad_norm": 0.0417444272173153,
      "learning_rate": 2.4418804056472227e-05,
      "loss": 0.3799,
      "step": 730
    },
    {
      "epoch": 1.3590449954086319,
      "grad_norm": 0.04538570278878064,
      "learning_rate": 2.3189159181485516e-05,
      "loss": 0.3921,
      "step": 740
    },
    {
      "epoch": 1.3774104683195592,
      "grad_norm": 0.043013131185681716,
      "learning_rate": 2.1981866606298683e-05,
      "loss": 0.3857,
      "step": 750
    },
    {
      "epoch": 1.3957759412304866,
      "grad_norm": 0.043552361092639345,
      "learning_rate": 2.079793285531618e-05,
      "loss": 0.3937,
      "step": 760
    },
    {
      "epoch": 1.4141414141414141,
      "grad_norm": 0.041014272519928036,
      "learning_rate": 1.963834497860192e-05,
      "loss": 0.3801,
      "step": 770
    },
    {
      "epoch": 1.4325068870523416,
      "grad_norm": 0.03865452597618009,
      "learning_rate": 1.8504069728972123e-05,
      "loss": 0.3834,
      "step": 780
    },
    {
      "epoch": 1.4508723599632691,
      "grad_norm": 0.04173720509030274,
      "learning_rate": 1.7396052756009574e-05,
      "loss": 0.3829,
      "step": 790
    },
    {
      "epoch": 1.4692378328741964,
      "grad_norm": 0.04256905712895654,
      "learning_rate": 1.631521781767214e-05,
      "loss": 0.3795,
      "step": 800
    },
    {
      "epoch": 1.487603305785124,
      "grad_norm": 0.0473389022063686,
      "learning_rate": 1.52624660101522e-05,
      "loss": 0.3801,
      "step": 810
    },
    {
      "epoch": 1.5059687786960514,
      "grad_norm": 0.0391194792265315,
      "learning_rate": 1.4238675016629338e-05,
      "loss": 0.3803,
      "step": 820
    },
    {
      "epoch": 1.5243342516069789,
      "grad_norm": 0.046767033474025464,
      "learning_rate": 1.3244698375542491e-05,
      "loss": 0.3963,
      "step": 830
    },
    {
      "epoch": 1.5426997245179064,
      "grad_norm": 0.042865074045739464,
      "learning_rate": 1.2281364768991804e-05,
      "loss": 0.3765,
      "step": 840
    },
    {
      "epoch": 1.5610651974288339,
      "grad_norm": 0.04190703980014869,
      "learning_rate": 1.134947733186315e-05,
      "loss": 0.3761,
      "step": 850
    },
    {
      "epoch": 1.5794306703397614,
      "grad_norm": 0.04183784386577142,
      "learning_rate": 1.0449812982251556e-05,
      "loss": 0.3852,
      "step": 860
    },
    {
      "epoch": 1.5977961432506889,
      "grad_norm": 0.03879386200844887,
      "learning_rate": 9.58312177374157e-06,
      "loss": 0.3775,
      "step": 870
    },
    {
      "epoch": 1.6161616161616161,
      "grad_norm": 0.0412465475003067,
      "learning_rate": 8.75012627008489e-06,
      "loss": 0.3818,
      "step": 880
    },
    {
      "epoch": 1.6345270890725436,
      "grad_norm": 0.04486626572785468,
      "learning_rate": 7.951520942796025e-06,
      "loss": 0.3871,
      "step": 890
    },
    {
      "epoch": 1.6528925619834711,
      "grad_norm": 0.040934222197984614,
      "learning_rate": 7.187971592168935e-06,
      "loss": 0.3739,
      "step": 900
    },
    {
      "epoch": 1.6712580348943984,
      "grad_norm": 0.0440521821621867,
      "learning_rate": 6.460114792196642e-06,
      "loss": 0.4001,
      "step": 910
    },
    {
      "epoch": 1.6896235078053259,
      "grad_norm": 0.04130542221023784,
      "learning_rate": 5.768557359857241e-06,
      "loss": 0.3833,
      "step": 920
    },
    {
      "epoch": 1.7079889807162534,
      "grad_norm": 0.041016879516671184,
      "learning_rate": 5.113875849208099e-06,
      "loss": 0.3871,
      "step": 930
    },
    {
      "epoch": 1.7263544536271809,
      "grad_norm": 0.04458130921484639,
      "learning_rate": 4.4966160707107076e-06,
      "loss": 0.3924,
      "step": 940
    },
    {
      "epoch": 1.7447199265381084,
      "grad_norm": 0.04483909598982454,
      "learning_rate": 3.917292636186332e-06,
      "loss": 0.3868,
      "step": 950
    },
    {
      "epoch": 1.7630853994490359,
      "grad_norm": 0.04374641202578716,
      "learning_rate": 3.376388529782215e-06,
      "loss": 0.3799,
      "step": 960
    },
    {
      "epoch": 1.7814508723599634,
      "grad_norm": 0.04643290880252759,
      "learning_rate": 2.8743547053058427e-06,
      "loss": 0.3873,
      "step": 970
    },
    {
      "epoch": 1.7998163452708908,
      "grad_norm": 0.04253533673018247,
      "learning_rate": 2.4116097102630907e-06,
      "loss": 0.3851,
      "step": 980
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.039915967047357025,
      "learning_rate": 1.9885393369134976e-06,
      "loss": 0.3773,
      "step": 990
    },
    {
      "epoch": 1.8365472910927456,
      "grad_norm": 0.04344253083810866,
      "learning_rate": 1.6054963006338742e-06,
      "loss": 0.3808,
      "step": 1000
    },
    {
      "epoch": 1.854912764003673,
      "grad_norm": 0.040517415128882496,
      "learning_rate": 1.2627999458580953e-06,
      "loss": 0.376,
      "step": 1010
    },
    {
      "epoch": 1.8732782369146006,
      "grad_norm": 0.040775411667955175,
      "learning_rate": 9.607359798384785e-07,
      "loss": 0.3824,
      "step": 1020
    },
    {
      "epoch": 1.8916437098255279,
      "grad_norm": 0.04533369281933566,
      "learning_rate": 6.995562344505214e-07,
      "loss": 0.3894,
      "step": 1030
    },
    {
      "epoch": 1.9100091827364554,
      "grad_norm": 0.0469790386931614,
      "learning_rate": 4.794784562397458e-07,
      "loss": 0.3746,
      "step": 1040
    },
    {
      "epoch": 1.9283746556473829,
      "grad_norm": 0.04179832849248068,
      "learning_rate": 3.006861248855408e-07,
      "loss": 0.3904,
      "step": 1050
    },
    {
      "epoch": 1.9467401285583104,
      "grad_norm": 0.04097669604600613,
      "learning_rate": 1.6332830023350064e-07,
      "loss": 0.3905,
      "step": 1060
    },
    {
      "epoch": 1.9651056014692379,
      "grad_norm": 0.043233457078553805,
      "learning_rate": 6.751949802362711e-08,
      "loss": 0.3777,
      "step": 1070
    },
    {
      "epoch": 1.9834710743801653,
      "grad_norm": 0.0432832073055713,
      "learning_rate": 1.3339594418138035e-08,
      "loss": 0.3719,
      "step": 1080
    },
    {
      "epoch": 1.9981634527089072,
      "step": 1088,
      "total_flos": 525248790528000.0,
      "train_loss": 0.4096706003388938,
      "train_runtime": 32183.0135,
      "train_samples_per_second": 4.058,
      "train_steps_per_second": 0.034
    }
  ],
  "logging_steps": 10,
  "max_steps": 1088,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 5000,
  "total_flos": 525248790528000.0,
  "train_batch_size": 10,
  "trial_name": null,
  "trial_params": null
}