{
  "best_metric": 0.028858734294772148,
  "best_model_checkpoint": "saves/psy-course/MentaLLaMA-chat-7B/train/fold4/checkpoint-1250",
  "epoch": 4.999522946283752,
  "eval_steps": 50,
  "global_step": 3275,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015265718919950386,
      "grad_norm": 1.6029620170593262,
      "learning_rate": 3.0487804878048782e-06,
      "loss": 1.7277,
      "step": 10
    },
    {
      "epoch": 0.030531437839900772,
      "grad_norm": 2.1076600551605225,
      "learning_rate": 6.0975609756097564e-06,
      "loss": 1.689,
      "step": 20
    },
    {
      "epoch": 0.04579715675985116,
      "grad_norm": 2.5429253578186035,
      "learning_rate": 9.146341463414634e-06,
      "loss": 1.6402,
      "step": 30
    },
    {
      "epoch": 0.061062875679801544,
      "grad_norm": 2.8307013511657715,
      "learning_rate": 1.2195121951219513e-05,
      "loss": 1.437,
      "step": 40
    },
    {
      "epoch": 0.07632859459975193,
      "grad_norm": 0.9449720978736877,
      "learning_rate": 1.524390243902439e-05,
      "loss": 0.796,
      "step": 50
    },
    {
      "epoch": 0.07632859459975193,
      "eval_loss": 0.614504337310791,
      "eval_runtime": 178.8075,
      "eval_samples_per_second": 6.515,
      "eval_steps_per_second": 6.515,
      "step": 50
    },
    {
      "epoch": 0.09159431351970232,
      "grad_norm": 1.0583168268203735,
      "learning_rate": 1.8292682926829268e-05,
      "loss": 0.5954,
      "step": 60
    },
    {
      "epoch": 0.10686003243965271,
      "grad_norm": 0.9122205972671509,
      "learning_rate": 2.134146341463415e-05,
      "loss": 0.4645,
      "step": 70
    },
    {
      "epoch": 0.12212575135960309,
      "grad_norm": 0.8755609393119812,
      "learning_rate": 2.4390243902439026e-05,
      "loss": 0.3116,
      "step": 80
    },
    {
      "epoch": 0.13739147027955348,
      "grad_norm": 0.8426439166069031,
      "learning_rate": 2.7439024390243906e-05,
      "loss": 0.1824,
      "step": 90
    },
    {
      "epoch": 0.15265718919950386,
      "grad_norm": 0.43453994393348694,
      "learning_rate": 3.048780487804878e-05,
      "loss": 0.1534,
      "step": 100
    },
    {
      "epoch": 0.15265718919950386,
      "eval_loss": 0.104978546500206,
      "eval_runtime": 178.8662,
      "eval_samples_per_second": 6.513,
      "eval_steps_per_second": 6.513,
      "step": 100
    },
    {
      "epoch": 0.16792290811945426,
      "grad_norm": 0.4413648843765259,
      "learning_rate": 3.353658536585366e-05,
      "loss": 0.1087,
      "step": 110
    },
    {
      "epoch": 0.18318862703940464,
      "grad_norm": 0.8687300682067871,
      "learning_rate": 3.6585365853658535e-05,
      "loss": 0.096,
      "step": 120
    },
    {
      "epoch": 0.19845434595935502,
      "grad_norm": 0.6343626379966736,
      "learning_rate": 3.9634146341463416e-05,
      "loss": 0.0853,
      "step": 130
    },
    {
      "epoch": 0.21372006487930542,
      "grad_norm": 0.7491863369941711,
      "learning_rate": 4.26829268292683e-05,
      "loss": 0.0767,
      "step": 140
    },
    {
      "epoch": 0.2289857837992558,
      "grad_norm": 0.495426744222641,
      "learning_rate": 4.573170731707318e-05,
      "loss": 0.0619,
      "step": 150
    },
    {
      "epoch": 0.2289857837992558,
      "eval_loss": 0.06294012069702148,
      "eval_runtime": 178.9812,
      "eval_samples_per_second": 6.509,
      "eval_steps_per_second": 6.509,
      "step": 150
    },
    {
      "epoch": 0.24425150271920618,
      "grad_norm": 0.7597019076347351,
      "learning_rate": 4.878048780487805e-05,
      "loss": 0.0873,
      "step": 160
    },
    {
      "epoch": 0.2595172216391566,
      "grad_norm": 0.804664671421051,
      "learning_rate": 5.182926829268293e-05,
      "loss": 0.0739,
      "step": 170
    },
    {
      "epoch": 0.27478294055910696,
      "grad_norm": 0.3810129165649414,
      "learning_rate": 5.487804878048781e-05,
      "loss": 0.0792,
      "step": 180
    },
    {
      "epoch": 0.29004865947905734,
      "grad_norm": 0.6029999852180481,
      "learning_rate": 5.792682926829268e-05,
      "loss": 0.0695,
      "step": 190
    },
    {
      "epoch": 0.3053143783990077,
      "grad_norm": 0.6588765382766724,
      "learning_rate": 6.097560975609756e-05,
      "loss": 0.0535,
      "step": 200
    },
    {
      "epoch": 0.3053143783990077,
      "eval_loss": 0.05500848591327667,
      "eval_runtime": 179.0405,
      "eval_samples_per_second": 6.507,
      "eval_steps_per_second": 6.507,
      "step": 200
    },
    {
      "epoch": 0.3205800973189581,
      "grad_norm": 0.5788701772689819,
      "learning_rate": 6.402439024390244e-05,
      "loss": 0.0593,
      "step": 210
    },
    {
      "epoch": 0.3358458162389085,
      "grad_norm": 0.7792841196060181,
      "learning_rate": 6.707317073170732e-05,
      "loss": 0.0566,
      "step": 220
    },
    {
      "epoch": 0.3511115351588589,
      "grad_norm": 0.7082734704017639,
      "learning_rate": 7.012195121951219e-05,
      "loss": 0.0449,
      "step": 230
    },
    {
      "epoch": 0.3663772540788093,
      "grad_norm": 0.5391858816146851,
      "learning_rate": 7.317073170731707e-05,
      "loss": 0.0636,
      "step": 240
    },
    {
      "epoch": 0.38164297299875966,
      "grad_norm": 0.47337889671325684,
      "learning_rate": 7.621951219512195e-05,
      "loss": 0.0576,
      "step": 250
    },
    {
      "epoch": 0.38164297299875966,
      "eval_loss": 0.04678180441260338,
      "eval_runtime": 178.9876,
      "eval_samples_per_second": 6.509,
      "eval_steps_per_second": 6.509,
      "step": 250
    },
    {
      "epoch": 0.39690869191871003,
      "grad_norm": 0.7037196159362793,
      "learning_rate": 7.926829268292683e-05,
      "loss": 0.0537,
      "step": 260
    },
    {
      "epoch": 0.4121744108386604,
      "grad_norm": 0.44716084003448486,
      "learning_rate": 8.231707317073171e-05,
      "loss": 0.0513,
      "step": 270
    },
    {
      "epoch": 0.42744012975861084,
      "grad_norm": 0.288339227437973,
      "learning_rate": 8.53658536585366e-05,
      "loss": 0.0524,
      "step": 280
    },
    {
      "epoch": 0.4427058486785612,
      "grad_norm": 0.5624875426292419,
      "learning_rate": 8.841463414634147e-05,
      "loss": 0.0575,
      "step": 290
    },
    {
      "epoch": 0.4579715675985116,
      "grad_norm": 0.3156270682811737,
      "learning_rate": 9.146341463414635e-05,
      "loss": 0.0501,
      "step": 300
    },
    {
      "epoch": 0.4579715675985116,
      "eval_loss": 0.046911776065826416,
      "eval_runtime": 178.9736,
      "eval_samples_per_second": 6.509,
      "eval_steps_per_second": 6.509,
      "step": 300
    },
    {
      "epoch": 0.473237286518462,
      "grad_norm": 0.5870501399040222,
      "learning_rate": 9.451219512195122e-05,
      "loss": 0.0536,
      "step": 310
    },
    {
      "epoch": 0.48850300543841235,
      "grad_norm": 0.27398020029067993,
      "learning_rate": 9.75609756097561e-05,
      "loss": 0.0369,
      "step": 320
    },
    {
      "epoch": 0.5037687243583627,
      "grad_norm": 0.6370400190353394,
      "learning_rate": 9.999988635788465e-05,
      "loss": 0.0474,
      "step": 330
    },
    {
      "epoch": 0.5190344432783132,
      "grad_norm": 0.6367189288139343,
      "learning_rate": 9.999590893808788e-05,
      "loss": 0.0455,
      "step": 340
    },
    {
      "epoch": 0.5343001621982635,
      "grad_norm": 0.47603678703308105,
      "learning_rate": 9.998624992909386e-05,
      "loss": 0.0456,
      "step": 350
    },
    {
      "epoch": 0.5343001621982635,
      "eval_loss": 0.042300350964069366,
      "eval_runtime": 178.9958,
      "eval_samples_per_second": 6.509,
      "eval_steps_per_second": 6.509,
      "step": 350
    },
    {
      "epoch": 0.5495658811182139,
      "grad_norm": 0.17798028886318207,
      "learning_rate": 9.997091042856284e-05,
      "loss": 0.0413,
      "step": 360
    },
    {
      "epoch": 0.5648316000381643,
      "grad_norm": 0.2773096561431885,
      "learning_rate": 9.994989217969224e-05,
      "loss": 0.0513,
      "step": 370
    },
    {
      "epoch": 0.5800973189581147,
      "grad_norm": 0.7576242685317993,
      "learning_rate": 9.992319757101863e-05,
      "loss": 0.0538,
      "step": 380
    },
    {
      "epoch": 0.5953630378780651,
      "grad_norm": 0.39998742938041687,
      "learning_rate": 9.98908296361462e-05,
      "loss": 0.0593,
      "step": 390
    },
    {
      "epoch": 0.6106287567980154,
      "grad_norm": 0.32975825667381287,
      "learning_rate": 9.98527920534021e-05,
      "loss": 0.0487,
      "step": 400
    },
    {
      "epoch": 0.6106287567980154,
      "eval_loss": 0.03924345597624779,
      "eval_runtime": 179.019,
      "eval_samples_per_second": 6.508,
      "eval_steps_per_second": 6.508,
      "step": 400
    },
    {
      "epoch": 0.6258944757179659,
      "grad_norm": 0.4215024709701538,
      "learning_rate": 9.980908914541844e-05,
      "loss": 0.0436,
      "step": 410
    },
    {
      "epoch": 0.6411601946379162,
      "grad_norm": 0.27699562907218933,
      "learning_rate": 9.975972587864095e-05,
      "loss": 0.0507,
      "step": 420
    },
    {
      "epoch": 0.6564259135578666,
      "grad_norm": 0.559887170791626,
      "learning_rate": 9.970470786276467e-05,
      "loss": 0.0378,
      "step": 430
    },
    {
      "epoch": 0.671691632477817,
      "grad_norm": 0.42969998717308044,
      "learning_rate": 9.964404135009648e-05,
      "loss": 0.0495,
      "step": 440
    },
    {
      "epoch": 0.6869573513977674,
      "grad_norm": 0.2953483462333679,
      "learning_rate": 9.957773323484454e-05,
      "loss": 0.0543,
      "step": 450
    },
    {
      "epoch": 0.6869573513977674,
      "eval_loss": 0.03886280581355095,
      "eval_runtime": 179.0976,
      "eval_samples_per_second": 6.505,
      "eval_steps_per_second": 6.505,
      "step": 450
    },
    {
      "epoch": 0.7022230703177178,
      "grad_norm": 0.2300979197025299,
      "learning_rate": 9.950579105233483e-05,
      "loss": 0.0386,
      "step": 460
    },
    {
      "epoch": 0.7174887892376681,
      "grad_norm": 0.2619139552116394,
      "learning_rate": 9.94282229781548e-05,
      "loss": 0.0402,
      "step": 470
    },
    {
      "epoch": 0.7327545081576186,
      "grad_norm": 0.49340057373046875,
      "learning_rate": 9.934503782722438e-05,
      "loss": 0.0444,
      "step": 480
    },
    {
      "epoch": 0.748020227077569,
      "grad_norm": 0.23474624752998352,
      "learning_rate": 9.925624505279411e-05,
      "loss": 0.0383,
      "step": 490
    },
    {
      "epoch": 0.7632859459975193,
      "grad_norm": 0.12562145292758942,
      "learning_rate": 9.916185474537098e-05,
      "loss": 0.0298,
      "step": 500
    },
    {
      "epoch": 0.7632859459975193,
      "eval_loss": 0.03485063835978508,
      "eval_runtime": 179.1055,
      "eval_samples_per_second": 6.505,
      "eval_steps_per_second": 6.505,
      "step": 500
    },
    {
      "epoch": 0.7785516649174697,
      "grad_norm": 0.33692803978919983,
      "learning_rate": 9.906187763157168e-05,
      "loss": 0.0472,
      "step": 510
    },
    {
      "epoch": 0.7938173838374201,
      "grad_norm": 0.28782883286476135,
      "learning_rate": 9.895632507290362e-05,
      "loss": 0.0479,
      "step": 520
    },
    {
      "epoch": 0.8090831027573705,
      "grad_norm": 0.2520039677619934,
      "learning_rate": 9.884520906447379e-05,
      "loss": 0.0359,
      "step": 530
    },
    {
      "epoch": 0.8243488216773208,
      "grad_norm": 0.3067379295825958,
      "learning_rate": 9.872854223362562e-05,
      "loss": 0.035,
      "step": 540
    },
    {
      "epoch": 0.8396145405972713,
      "grad_norm": 0.5226966738700867,
      "learning_rate": 9.860633783850406e-05,
      "loss": 0.0378,
      "step": 550
    },
    {
      "epoch": 0.8396145405972713,
      "eval_loss": 0.03473576903343201,
      "eval_runtime": 179.0927,
      "eval_samples_per_second": 6.505,
      "eval_steps_per_second": 6.505,
      "step": 550
    },
    {
      "epoch": 0.8548802595172217,
      "grad_norm": 0.16309522092342377,
      "learning_rate": 9.847860976654879e-05,
      "loss": 0.0404,
      "step": 560
    },
    {
      "epoch": 0.870145978437172,
      "grad_norm": 0.18132278323173523,
      "learning_rate": 9.834537253291616e-05,
      "loss": 0.0401,
      "step": 570
    },
    {
      "epoch": 0.8854116973571224,
      "grad_norm": 0.583713948726654,
      "learning_rate": 9.820664127882957e-05,
      "loss": 0.0468,
      "step": 580
    },
    {
      "epoch": 0.9006774162770728,
      "grad_norm": 0.3539969325065613,
      "learning_rate": 9.806243176985888e-05,
      "loss": 0.0437,
      "step": 590
    },
    {
      "epoch": 0.9159431351970232,
      "grad_norm": 0.3411082923412323,
      "learning_rate": 9.791276039412875e-05,
      "loss": 0.0387,
      "step": 600
    },
    {
      "epoch": 0.9159431351970232,
      "eval_loss": 0.032861627638339996,
      "eval_runtime": 179.1835,
      "eval_samples_per_second": 6.502,
      "eval_steps_per_second": 6.502,
      "step": 600
    },
    {
      "epoch": 0.9312088541169735,
      "grad_norm": 0.22567597031593323,
      "learning_rate": 9.775764416045628e-05,
      "loss": 0.0502,
      "step": 610
    },
    {
      "epoch": 0.946474573036924,
      "grad_norm": 0.29538026452064514,
      "learning_rate": 9.759710069641814e-05,
      "loss": 0.042,
      "step": 620
    },
    {
      "epoch": 0.9617402919568744,
      "grad_norm": 0.20513112843036652,
      "learning_rate": 9.743114824634734e-05,
      "loss": 0.0283,
      "step": 630
    },
    {
      "epoch": 0.9770060108768247,
      "grad_norm": 0.26253455877304077,
      "learning_rate": 9.725980566925989e-05,
      "loss": 0.0367,
      "step": 640
    },
    {
      "epoch": 0.9922717297967751,
      "grad_norm": 0.2190350592136383,
      "learning_rate": 9.708309243671165e-05,
      "loss": 0.0275,
      "step": 650
    },
    {
      "epoch": 0.9922717297967751,
      "eval_loss": 0.035241108387708664,
      "eval_runtime": 179.2823,
      "eval_samples_per_second": 6.498,
      "eval_steps_per_second": 6.498,
      "step": 650
    },
    {
      "epoch": 1.0075374487167255,
      "grad_norm": 0.09830069541931152,
      "learning_rate": 9.690102863058563e-05,
      "loss": 0.0321,
      "step": 660
    },
    {
      "epoch": 1.0228031676366758,
      "grad_norm": 0.2949357032775879,
      "learning_rate": 9.67136349408098e-05,
      "loss": 0.0293,
      "step": 670
    },
    {
      "epoch": 1.0380688865566263,
      "grad_norm": 0.10271728783845901,
      "learning_rate": 9.652093266300583e-05,
      "loss": 0.0309,
      "step": 680
    },
    {
      "epoch": 1.0533346054765766,
      "grad_norm": 0.2911938726902008,
      "learning_rate": 9.632294369606916e-05,
      "loss": 0.0213,
      "step": 690
    },
    {
      "epoch": 1.068600324396527,
      "grad_norm": 0.23335903882980347,
      "learning_rate": 9.61196905396802e-05,
      "loss": 0.0348,
      "step": 700
    },
    {
      "epoch": 1.068600324396527,
      "eval_loss": 0.03389936685562134,
      "eval_runtime": 179.1248,
      "eval_samples_per_second": 6.504,
      "eval_steps_per_second": 6.504,
      "step": 700
    },
    {
      "epoch": 1.0838660433164775,
      "grad_norm": 0.26263928413391113,
      "learning_rate": 9.591119629174764e-05,
      "loss": 0.0398,
      "step": 710
    },
    {
      "epoch": 1.0991317622364278,
      "grad_norm": 0.15275321900844574,
      "learning_rate": 9.569748464578343e-05,
      "loss": 0.0229,
      "step": 720
    },
    {
      "epoch": 1.1143974811563782,
      "grad_norm": 0.23484571278095245,
      "learning_rate": 9.54785798882103e-05,
      "loss": 0.029,
      "step": 730
    },
    {
      "epoch": 1.1296632000763287,
      "grad_norm": 0.18005479872226715,
      "learning_rate": 9.525450689560181e-05,
      "loss": 0.0266,
      "step": 740
    },
    {
      "epoch": 1.144928918996279,
      "grad_norm": 0.17082622647285461,
      "learning_rate": 9.502529113185532e-05,
      "loss": 0.0346,
      "step": 750
    },
    {
      "epoch": 1.144928918996279,
      "eval_loss": 0.03384435176849365,
      "eval_runtime": 179.1418,
      "eval_samples_per_second": 6.503,
      "eval_steps_per_second": 6.503,
      "step": 750
    },
    {
      "epoch": 1.1601946379162293,
      "grad_norm": 0.31557226181030273,
      "learning_rate": 9.479095864529828e-05,
      "loss": 0.0297,
      "step": 760
    },
    {
      "epoch": 1.1754603568361797,
      "grad_norm": 0.17261050641536713,
      "learning_rate": 9.455153606572806e-05,
      "loss": 0.0242,
      "step": 770
    },
    {
      "epoch": 1.1907260757561302,
      "grad_norm": 0.25501230359077454,
      "learning_rate": 9.430705060138569e-05,
      "loss": 0.0325,
      "step": 780
    },
    {
      "epoch": 1.2059917946760805,
      "grad_norm": 0.47437065839767456,
      "learning_rate": 9.405753003586395e-05,
      "loss": 0.0288,
      "step": 790
    },
    {
      "epoch": 1.2212575135960309,
      "grad_norm": 0.20840288698673248,
      "learning_rate": 9.38030027249499e-05,
      "loss": 0.0321,
      "step": 800
    },
    {
      "epoch": 1.2212575135960309,
      "eval_loss": 0.033740561455488205,
      "eval_runtime": 179.1979,
      "eval_samples_per_second": 6.501,
      "eval_steps_per_second": 6.501,
      "step": 800
    },
    {
      "epoch": 1.2365232325159814,
      "grad_norm": 0.24225717782974243,
      "learning_rate": 9.354349759340263e-05,
      "loss": 0.0278,
      "step": 810
    },
    {
      "epoch": 1.2517889514359317,
      "grad_norm": 0.20259130001068115,
      "learning_rate": 9.327904413166615e-05,
      "loss": 0.0291,
      "step": 820
    },
    {
      "epoch": 1.267054670355882,
      "grad_norm": 0.4767257273197174,
      "learning_rate": 9.300967239251798e-05,
      "loss": 0.0374,
      "step": 830
    },
    {
      "epoch": 1.2823203892758324,
      "grad_norm": 0.20219506323337555,
      "learning_rate": 9.27354129876541e-05,
      "loss": 0.03,
      "step": 840
    },
    {
      "epoch": 1.297586108195783,
      "grad_norm": 0.19528479874134064,
      "learning_rate": 9.245629708421008e-05,
      "loss": 0.0288,
      "step": 850
    },
    {
      "epoch": 1.297586108195783,
      "eval_loss": 0.0333208329975605,
      "eval_runtime": 179.2291,
      "eval_samples_per_second": 6.5,
      "eval_steps_per_second": 6.5,
      "step": 850
    },
    {
      "epoch": 1.3128518271157332,
      "grad_norm": 0.3612363934516907,
      "learning_rate": 9.217235640121926e-05,
      "loss": 0.0311,
      "step": 860
    },
    {
      "epoch": 1.3281175460356835,
      "grad_norm": 0.4072886109352112,
      "learning_rate": 9.188362320600812e-05,
      "loss": 0.0371,
      "step": 870
    },
    {
      "epoch": 1.343383264955634,
      "grad_norm": 0.2643483281135559,
      "learning_rate": 9.159013031052943e-05,
      "loss": 0.0266,
      "step": 880
    },
    {
      "epoch": 1.3586489838755844,
      "grad_norm": 0.30413269996643066,
      "learning_rate": 9.129191106763346e-05,
      "loss": 0.0276,
      "step": 890
    },
    {
      "epoch": 1.3739147027955347,
      "grad_norm": 0.23679038882255554,
      "learning_rate": 9.098899936727771e-05,
      "loss": 0.028,
      "step": 900
    },
    {
      "epoch": 1.3739147027955347,
      "eval_loss": 0.03594866767525673,
      "eval_runtime": 179.217,
      "eval_samples_per_second": 6.501,
      "eval_steps_per_second": 6.501,
      "step": 900
    },
    {
      "epoch": 1.3891804217154853,
      "grad_norm": 0.4948261082172394,
      "learning_rate": 9.068142963267558e-05,
      "loss": 0.0262,
      "step": 910
    },
    {
      "epoch": 1.4044461406354356,
      "grad_norm": 0.27103114128112793,
      "learning_rate": 9.036923681638463e-05,
      "loss": 0.0291,
      "step": 920
    },
    {
      "epoch": 1.419711859555386,
      "grad_norm": 0.17580902576446533,
      "learning_rate": 9.00524563963343e-05,
      "loss": 0.0205,
      "step": 930
    },
    {
      "epoch": 1.4349775784753362,
      "grad_norm": 0.22460690140724182,
      "learning_rate": 8.973112437179436e-05,
      "loss": 0.033,
      "step": 940
    },
    {
      "epoch": 1.4502432973952868,
      "grad_norm": 0.16176696121692657,
      "learning_rate": 8.940527725928383e-05,
      "loss": 0.0277,
      "step": 950
    },
    {
      "epoch": 1.4502432973952868,
      "eval_loss": 0.03296082094311714,
      "eval_runtime": 179.256,
      "eval_samples_per_second": 6.499,
      "eval_steps_per_second": 6.499,
      "step": 950
    },
    {
      "epoch": 1.4655090163152371,
      "grad_norm": 0.16683737933635712,
      "learning_rate": 8.90749520884212e-05,
      "loss": 0.028,
      "step": 960
    },
    {
      "epoch": 1.4807747352351874,
      "grad_norm": 0.23072558641433716,
      "learning_rate": 8.874018639771637e-05,
      "loss": 0.031,
      "step": 970
    },
    {
      "epoch": 1.4960404541551378,
      "grad_norm": 0.18204113841056824,
      "learning_rate": 8.840101823030471e-05,
      "loss": 0.0261,
      "step": 980
    },
    {
      "epoch": 1.5113061730750883,
      "grad_norm": 0.2924092411994934,
      "learning_rate": 8.805748612962382e-05,
      "loss": 0.0295,
      "step": 990
    },
    {
      "epoch": 1.5265718919950386,
      "grad_norm": 0.1391419917345047,
      "learning_rate": 8.77096291350334e-05,
      "loss": 0.0202,
      "step": 1000
    },
    {
      "epoch": 1.5265718919950386,
      "eval_loss": 0.03339965641498566,
      "eval_runtime": 179.5613,
      "eval_samples_per_second": 6.488,
      "eval_steps_per_second": 6.488,
      "step": 1000
    },
    {
      "epoch": 1.541837610914989,
      "grad_norm": 0.2946327030658722,
      "learning_rate": 8.735748677737874e-05,
      "loss": 0.0299,
      "step": 1010
    },
    {
      "epoch": 1.5571033298349395,
      "grad_norm": 0.8059619069099426,
      "learning_rate": 8.700109907449845e-05,
      "loss": 0.0223,
      "step": 1020
    },
    {
      "epoch": 1.5723690487548898,
      "grad_norm": 0.09737138450145721,
      "learning_rate": 8.66405065266768e-05,
      "loss": 0.0251,
      "step": 1030
    },
    {
      "epoch": 1.5876347676748401,
      "grad_norm": 0.20474199950695038,
      "learning_rate": 8.627575011204115e-05,
      "loss": 0.0241,
      "step": 1040
    },
    {
      "epoch": 1.6029004865947907,
      "grad_norm": 0.13863246142864227,
      "learning_rate": 8.590687128190516e-05,
      "loss": 0.0291,
      "step": 1050
    },
    {
      "epoch": 1.6029004865947907,
      "eval_loss": 0.03433850780129433,
      "eval_runtime": 179.7512,
      "eval_samples_per_second": 6.481,
      "eval_steps_per_second": 6.481,
      "step": 1050
    },
    {
      "epoch": 1.618166205514741,
      "grad_norm": 0.12294421344995499,
      "learning_rate": 8.553391195605833e-05,
      "loss": 0.0247,
      "step": 1060
    },
    {
      "epoch": 1.6334319244346913,
      "grad_norm": 0.2797018587589264,
      "learning_rate": 8.515691451800205e-05,
      "loss": 0.0426,
      "step": 1070
    },
    {
      "epoch": 1.6486976433546419,
      "grad_norm": 0.19925975799560547,
      "learning_rate": 8.477592181013316e-05,
      "loss": 0.0237,
      "step": 1080
    },
    {
      "epoch": 1.663963362274592,
      "grad_norm": 0.16685841977596283,
      "learning_rate": 8.439097712887531e-05,
      "loss": 0.0274,
      "step": 1090
    },
    {
      "epoch": 1.6792290811945425,
      "grad_norm": 0.2913149893283844,
      "learning_rate": 8.400212421975865e-05,
      "loss": 0.0402,
      "step": 1100
    },
    {
      "epoch": 1.6792290811945425,
      "eval_loss": 0.03206111863255501,
      "eval_runtime": 179.642,
      "eval_samples_per_second": 6.485,
      "eval_steps_per_second": 6.485,
      "step": 1100
    },
    {
      "epoch": 1.694494800114493,
      "grad_norm": 0.19945311546325684,
      "learning_rate": 8.360940727244859e-05,
      "loss": 0.0343,
      "step": 1110
    },
    {
      "epoch": 1.7097605190344431,
      "grad_norm": 0.20246616005897522,
      "learning_rate": 8.321287091572403e-05,
      "loss": 0.0233,
      "step": 1120
    },
    {
      "epoch": 1.7250262379543937,
      "grad_norm": 0.455074667930603,
      "learning_rate": 8.281256021240566e-05,
      "loss": 0.0283,
      "step": 1130
    },
    {
      "epoch": 1.740291956874344,
      "grad_norm": 0.1479114294052124,
      "learning_rate": 8.240852065423506e-05,
      "loss": 0.0351,
      "step": 1140
    },
    {
      "epoch": 1.7555576757942943,
      "grad_norm": 0.3095378577709198,
      "learning_rate": 8.20007981567048e-05,
      "loss": 0.0268,
      "step": 1150
    },
    {
      "epoch": 1.7555576757942943,
      "eval_loss": 0.03280867636203766,
      "eval_runtime": 179.7524,
      "eval_samples_per_second": 6.481,
      "eval_steps_per_second": 6.481,
      "step": 1150
    },
    {
      "epoch": 1.7708233947142449,
      "grad_norm": 0.1850660890340805,
      "learning_rate": 8.158943905384082e-05,
      "loss": 0.0322,
      "step": 1160
    },
    {
      "epoch": 1.7860891136341952,
      "grad_norm": 0.23737244307994843,
      "learning_rate": 8.117449009293668e-05,
      "loss": 0.0291,
      "step": 1170
    },
    {
      "epoch": 1.8013548325541455,
      "grad_norm": 0.18103042244911194,
      "learning_rate": 8.075599842924139e-05,
      "loss": 0.0276,
      "step": 1180
    },
    {
      "epoch": 1.816620551474096,
      "grad_norm": 0.18441970646381378,
      "learning_rate": 8.033401162060049e-05,
      "loss": 0.0343,
      "step": 1190
    },
    {
      "epoch": 1.8318862703940464,
      "grad_norm": 0.36745738983154297,
      "learning_rate": 7.990857762205157e-05,
      "loss": 0.0411,
      "step": 1200
    },
    {
      "epoch": 1.8318862703940464,
      "eval_loss": 0.032041117548942566,
      "eval_runtime": 179.8085,
      "eval_samples_per_second": 6.479,
      "eval_steps_per_second": 6.479,
      "step": 1200
    },
    {
      "epoch": 1.8471519893139967,
      "grad_norm": 0.28123798966407776,
      "learning_rate": 7.947974478037468e-05,
      "loss": 0.0355,
      "step": 1210
    },
    {
      "epoch": 1.8624177082339473,
      "grad_norm": 0.6014788150787354,
      "learning_rate": 7.904756182859797e-05,
      "loss": 0.0235,
      "step": 1220
    },
    {
      "epoch": 1.8776834271538976,
      "grad_norm": 0.11423056572675705,
      "learning_rate": 7.861207788045984e-05,
      "loss": 0.0327,
      "step": 1230
    },
    {
      "epoch": 1.892949146073848,
      "grad_norm": 0.33343955874443054,
      "learning_rate": 7.817334242482738e-05,
      "loss": 0.028,
      "step": 1240
    },
    {
      "epoch": 1.9082148649937984,
      "grad_norm": 0.12100326269865036,
      "learning_rate": 7.773140532007262e-05,
      "loss": 0.0262,
      "step": 1250
    },
    {
      "epoch": 1.9082148649937984,
      "eval_loss": 0.028858734294772148,
      "eval_runtime": 179.691,
      "eval_samples_per_second": 6.483,
      "eval_steps_per_second": 6.483,
      "step": 1250
    },
    {
      "epoch": 1.9234805839137485,
      "grad_norm": 0.1283182054758072,
      "learning_rate": 7.728631678840638e-05,
      "loss": 0.0267,
      "step": 1260
    },
    {
      "epoch": 1.938746302833699,
      "grad_norm": 0.11831828206777573,
      "learning_rate": 7.683812741017112e-05,
      "loss": 0.0252,
      "step": 1270
    },
    {
      "epoch": 1.9540120217536494,
      "grad_norm": 0.33695539832115173,
      "learning_rate": 7.638688811809274e-05,
      "loss": 0.0294,
      "step": 1280
    },
    {
      "epoch": 1.9692777406735997,
      "grad_norm": 0.19394966959953308,
      "learning_rate": 7.593265019149275e-05,
      "loss": 0.029,
      "step": 1290
    },
    {
      "epoch": 1.9845434595935503,
      "grad_norm": 0.24795125424861908,
      "learning_rate": 7.547546525046073e-05,
      "loss": 0.0318,
      "step": 1300
    },
    {
      "epoch": 1.9845434595935503,
      "eval_loss": 0.02917238511145115,
      "eval_runtime": 179.7965,
      "eval_samples_per_second": 6.48,
      "eval_steps_per_second": 6.48,
      "step": 1300
    },
    {
      "epoch": 1.9998091785135006,
      "grad_norm": 0.08313370496034622,
      "learning_rate": 7.501538524998812e-05,
      "loss": 0.0305,
      "step": 1310
    },
    {
      "epoch": 2.015074897433451,
      "grad_norm": 0.16452887654304504,
      "learning_rate": 7.455246247406406e-05,
      "loss": 0.0137,
      "step": 1320
    },
    {
      "epoch": 2.0303406163534015,
      "grad_norm": 0.26041746139526367,
      "learning_rate": 7.408674952973382e-05,
      "loss": 0.0173,
      "step": 1330
    },
    {
      "epoch": 2.0456063352733516,
      "grad_norm": 0.19397439062595367,
      "learning_rate": 7.361829934112036e-05,
      "loss": 0.0217,
      "step": 1340
    },
    {
      "epoch": 2.060872054193302,
      "grad_norm": 0.19004268944263458,
      "learning_rate": 7.314716514341006e-05,
      "loss": 0.0159,
      "step": 1350
    },
    {
      "epoch": 2.060872054193302,
      "eval_loss": 0.030632687732577324,
      "eval_runtime": 179.7659,
      "eval_samples_per_second": 6.481,
      "eval_steps_per_second": 6.481,
      "step": 1350
    },
    {
      "epoch": 2.0761377731132526,
      "grad_norm": 0.2005777806043625,
      "learning_rate": 7.267340047680305e-05,
      "loss": 0.0168,
      "step": 1360
    },
    {
      "epoch": 2.0914034920332027,
      "grad_norm": 0.2647518813610077,
      "learning_rate": 7.21970591804287e-05,
      "loss": 0.0159,
      "step": 1370
    },
    {
      "epoch": 2.1066692109531533,
      "grad_norm": 0.27719810605049133,
      "learning_rate": 7.171819538622747e-05,
      "loss": 0.0202,
      "step": 1380
    },
    {
      "epoch": 2.121934929873104,
      "grad_norm": 0.17780855298042297,
      "learning_rate": 7.123686351279914e-05,
      "loss": 0.0182,
      "step": 1390
    },
    {
      "epoch": 2.137200648793054,
      "grad_norm": 0.2095697820186615,
      "learning_rate": 7.07531182592187e-05,
      "loss": 0.019,
      "step": 1400
    },
    {
      "epoch": 2.137200648793054,
      "eval_loss": 0.031268153339624405,
      "eval_runtime": 179.7579,
      "eval_samples_per_second": 6.481,
      "eval_steps_per_second": 6.481,
      "step": 1400
    },
    {
      "epoch": 2.1524663677130045,
      "grad_norm": 0.3520405888557434,
      "learning_rate": 7.026701459882026e-05,
      "loss": 0.0271,
      "step": 1410
    },
    {
      "epoch": 2.167732086632955,
      "grad_norm": 0.16760867834091187,
      "learning_rate": 6.977860777294988e-05,
      "loss": 0.0193,
      "step": 1420
    },
    {
      "epoch": 2.182997805552905,
      "grad_norm": 0.1556943655014038,
      "learning_rate": 6.92879532846878e-05,
      "loss": 0.017,
      "step": 1430
    },
    {
      "epoch": 2.1982635244728557,
      "grad_norm": 0.17596006393432617,
      "learning_rate": 6.879510689254104e-05,
      "loss": 0.016,
      "step": 1440
    },
    {
      "epoch": 2.213529243392806,
      "grad_norm": 0.18705613911151886,
      "learning_rate": 6.830012460410697e-05,
      "loss": 0.022,
      "step": 1450
    },
    {
      "epoch": 2.213529243392806,
      "eval_loss": 0.03142455592751503,
      "eval_runtime": 179.7594,
      "eval_samples_per_second": 6.481,
      "eval_steps_per_second": 6.481,
      "step": 1450
    },
    {
      "epoch": 2.2287949623127563,
      "grad_norm": 0.22514861822128296,
      "learning_rate": 6.780306266970851e-05,
      "loss": 0.0143,
      "step": 1460
    },
    {
      "epoch": 2.244060681232707,
      "grad_norm": 0.3207738995552063,
      "learning_rate": 6.73039775760018e-05,
      "loss": 0.0203,
      "step": 1470
    },
    {
      "epoch": 2.2593264001526574,
      "grad_norm": 0.14879602193832397,
      "learning_rate": 6.680292603955702e-05,
      "loss": 0.0153,
      "step": 1480
    },
    {
      "epoch": 2.2745921190726075,
      "grad_norm": 0.16453400254249573,
      "learning_rate": 6.629996500041299e-05,
      "loss": 0.0181,
      "step": 1490
    },
    {
      "epoch": 2.289857837992558,
      "grad_norm": 0.23307697474956512,
      "learning_rate": 6.579515161560649e-05,
      "loss": 0.0133,
      "step": 1500
    },
    {
      "epoch": 2.289857837992558,
      "eval_loss": 0.032863009721040726,
      "eval_runtime": 179.7521,
      "eval_samples_per_second": 6.481,
      "eval_steps_per_second": 6.481,
      "step": 1500
    },
    {
      "epoch": 2.305123556912508,
      "grad_norm": 0.29715099930763245,
      "learning_rate": 6.528854325267692e-05,
      "loss": 0.0225,
      "step": 1510
    },
    {
      "epoch": 2.3203892758324587,
      "grad_norm": 0.27671730518341064,
      "learning_rate": 6.478019748314686e-05,
      "loss": 0.0193,
      "step": 1520
    },
    {
      "epoch": 2.3356549947524092,
      "grad_norm": 0.08666820079088211,
      "learning_rate": 6.42701720759797e-05,
      "loss": 0.0206,
      "step": 1530
    },
    {
      "epoch": 2.3509207136723593,
      "grad_norm": 0.08629153668880463,
      "learning_rate": 6.375852499101467e-05,
      "loss": 0.0147,
      "step": 1540
    },
    {
      "epoch": 2.36618643259231,
      "grad_norm": 0.11571142822504044,
      "learning_rate": 6.324531437238019e-05,
      "loss": 0.0134,
      "step": 1550
    },
    {
      "epoch": 2.36618643259231,
      "eval_loss": 0.031696803867816925,
      "eval_runtime": 179.7802,
      "eval_samples_per_second": 6.48,
      "eval_steps_per_second": 6.48,
      "step": 1550
    },
    {
      "epoch": 2.3814521515122604,
      "grad_norm": 0.2257205992937088,
      "learning_rate": 6.273059854188636e-05,
      "loss": 0.0228,
      "step": 1560
    },
    {
      "epoch": 2.3967178704322105,
      "grad_norm": 0.44024819135665894,
      "learning_rate": 6.221443599239721e-05,
      "loss": 0.0203,
      "step": 1570
    },
    {
      "epoch": 2.411983589352161,
      "grad_norm": 0.20170512795448303,
      "learning_rate": 6.169688538118342e-05,
      "loss": 0.0227,
      "step": 1580
    },
    {
      "epoch": 2.4272493082721116,
      "grad_norm": 0.15628792345523834,
      "learning_rate": 6.117800552325655e-05,
      "loss": 0.0181,
      "step": 1590
    },
    {
      "epoch": 2.4425150271920617,
      "grad_norm": 0.17330731451511383,
      "learning_rate": 6.06578553846852e-05,
      "loss": 0.0185,
      "step": 1600
    },
    {
      "epoch": 2.4425150271920617,
      "eval_loss": 0.029930831864476204,
      "eval_runtime": 179.7394,
      "eval_samples_per_second": 6.482,
      "eval_steps_per_second": 6.482,
      "step": 1600
    },
    {
      "epoch": 2.4577807461120122,
      "grad_norm": 0.23071150481700897,
      "learning_rate": 6.013649407589401e-05,
      "loss": 0.0236,
      "step": 1610
    },
    {
      "epoch": 2.473046465031963,
      "grad_norm": 0.17635267972946167,
      "learning_rate": 5.961398084494634e-05,
      "loss": 0.0211,
      "step": 1620
    },
    {
      "epoch": 2.488312183951913,
      "grad_norm": 0.2662081718444824,
      "learning_rate": 5.909037507081121e-05,
      "loss": 0.0247,
      "step": 1630
    },
    {
      "epoch": 2.5035779028718634,
      "grad_norm": 0.16230575740337372,
      "learning_rate": 5.8565736256615434e-05,
      "loss": 0.0243,
      "step": 1640
    },
    {
      "epoch": 2.518843621791814,
      "grad_norm": 0.11765624582767487,
      "learning_rate": 5.8040124022881625e-05,
      "loss": 0.0229,
      "step": 1650
    },
    {
      "epoch": 2.518843621791814,
      "eval_loss": 0.030760467052459717,
      "eval_runtime": 179.8009,
      "eval_samples_per_second": 6.479,
      "eval_steps_per_second": 6.479,
      "step": 1650
    },
    {
      "epoch": 2.534109340711764,
      "grad_norm": 0.2117856740951538,
      "learning_rate": 5.751359810075284e-05,
      "loss": 0.0155,
      "step": 1660
    },
    {
      "epoch": 2.5493750596317146,
      "grad_norm": 0.1411379873752594,
      "learning_rate": 5.6986218325204676e-05,
      "loss": 0.0133,
      "step": 1670
    },
    {
      "epoch": 2.5646407785516647,
      "grad_norm": 0.1770981103181839,
      "learning_rate": 5.645804462824556e-05,
      "loss": 0.0159,
      "step": 1680
    },
    {
      "epoch": 2.5799064974716153,
      "grad_norm": 0.3099724352359772,
      "learning_rate": 5.5929137032106005e-05,
      "loss": 0.0182,
      "step": 1690
    },
    {
      "epoch": 2.595172216391566,
      "grad_norm": 0.22490672767162323,
      "learning_rate": 5.53995556424176e-05,
      "loss": 0.0224,
      "step": 1700
    },
    {
      "epoch": 2.595172216391566,
      "eval_loss": 0.030679209157824516,
      "eval_runtime": 179.8574,
      "eval_samples_per_second": 6.477,
      "eval_steps_per_second": 6.477,
      "step": 1700
    },
    {
      "epoch": 2.610437935311516,
      "grad_norm": 0.13174369931221008,
      "learning_rate": 5.4869360641382615e-05,
      "loss": 0.0152,
      "step": 1710
    },
    {
      "epoch": 2.6257036542314665,
      "grad_norm": 0.21557074785232544,
      "learning_rate": 5.433861228093471e-05,
      "loss": 0.0209,
      "step": 1720
    },
    {
      "epoch": 2.640969373151417,
      "grad_norm": 0.13589583337306976,
      "learning_rate": 5.380737087589197e-05,
      "loss": 0.0144,
      "step": 1730
    },
    {
      "epoch": 2.656235092071367,
      "grad_norm": 0.23737147450447083,
      "learning_rate": 5.327569679710256e-05,
      "loss": 0.0236,
      "step": 1740
    },
    {
      "epoch": 2.6715008109913176,
      "grad_norm": 0.15822412073612213,
      "learning_rate": 5.274365046458416e-05,
      "loss": 0.0159,
      "step": 1750
    },
    {
      "epoch": 2.6715008109913176,
      "eval_loss": 0.03090774454176426,
      "eval_runtime": 179.9376,
      "eval_samples_per_second": 6.474,
      "eval_steps_per_second": 6.474,
      "step": 1750
    },
    {
      "epoch": 2.686766529911268,
      "grad_norm": 0.31121695041656494,
      "learning_rate": 5.2211292340657804e-05,
      "loss": 0.0209,
      "step": 1760
    },
    {
      "epoch": 2.7020322488312183,
      "grad_norm": 0.1423243284225464,
      "learning_rate": 5.167868292307678e-05,
      "loss": 0.0225,
      "step": 1770
    },
    {
      "epoch": 2.717297967751169,
      "grad_norm": 0.18490996956825256,
      "learning_rate": 5.114588273815173e-05,
      "loss": 0.019,
      "step": 1780
    },
    {
      "epoch": 2.732563686671119,
      "grad_norm": 0.2611638307571411,
      "learning_rate": 5.061295233387223e-05,
      "loss": 0.0228,
      "step": 1790
    },
    {
      "epoch": 2.7478294055910695,
      "grad_norm": 0.15606072545051575,
      "learning_rate": 5.007995227302617e-05,
      "loss": 0.0185,
      "step": 1800
    },
    {
      "epoch": 2.7478294055910695,
      "eval_loss": 0.03144565969705582,
      "eval_runtime": 179.8019,
      "eval_samples_per_second": 6.479,
      "eval_steps_per_second": 6.479,
      "step": 1800
    },
    {
      "epoch": 2.76309512451102,
      "grad_norm": 0.09613101929426193,
      "learning_rate": 4.954694312631729e-05,
      "loss": 0.0176,
      "step": 1810
    },
    {
      "epoch": 2.7783608434309706,
      "grad_norm": 0.23314177989959717,
      "learning_rate": 4.901398546548181e-05,
      "loss": 0.0179,
      "step": 1820
    },
    {
      "epoch": 2.7936265623509207,
      "grad_norm": 0.2562658488750458,
      "learning_rate": 4.848113985640513e-05,
      "loss": 0.0163,
      "step": 1830
    },
    {
      "epoch": 2.808892281270871,
      "grad_norm": 0.13945740461349487,
      "learning_rate": 4.794846685223884e-05,
      "loss": 0.0194,
      "step": 1840
    },
    {
      "epoch": 2.8241580001908213,
      "grad_norm": 0.2183397263288498,
      "learning_rate": 4.741602698651966e-05,
      "loss": 0.0203,
      "step": 1850
    },
    {
      "epoch": 2.8241580001908213,
      "eval_loss": 0.03035680390894413,
      "eval_runtime": 179.8704,
      "eval_samples_per_second": 6.477,
      "eval_steps_per_second": 6.477,
      "step": 1850
    },
    {
      "epoch": 2.839423719110772,
      "grad_norm": 0.10274723172187805,
      "learning_rate": 4.6883880766290086e-05,
      "loss": 0.0208,
      "step": 1860
    },
    {
      "epoch": 2.8546894380307224,
      "grad_norm": 0.1990806758403778,
      "learning_rate": 4.635208866522251e-05,
      "loss": 0.0149,
      "step": 1870
    },
    {
      "epoch": 2.8699551569506725,
      "grad_norm": 0.15856200456619263,
      "learning_rate": 4.5820711116746785e-05,
      "loss": 0.0134,
      "step": 1880
    },
    {
      "epoch": 2.885220875870623,
      "grad_norm": 0.30222058296203613,
      "learning_rate": 4.528980850718255e-05,
      "loss": 0.0216,
      "step": 1890
    },
    {
      "epoch": 2.9004865947905736,
      "grad_norm": 0.13257752358913422,
      "learning_rate": 4.475944116887695e-05,
      "loss": 0.0159,
      "step": 1900
    },
    {
      "epoch": 2.9004865947905736,
      "eval_loss": 0.031748704612255096,
      "eval_runtime": 179.7821,
      "eval_samples_per_second": 6.48,
      "eval_steps_per_second": 6.48,
      "step": 1900
    },
    {
      "epoch": 2.9157523137105237,
      "grad_norm": 0.2612052261829376,
      "learning_rate": 4.4229669373348226e-05,
      "loss": 0.0215,
      "step": 1910
    },
    {
      "epoch": 2.9310180326304742,
      "grad_norm": 0.15642981231212616,
      "learning_rate": 4.3700553324436575e-05,
      "loss": 0.0191,
      "step": 1920
    },
    {
      "epoch": 2.9462837515504248,
      "grad_norm": 0.24419263005256653,
      "learning_rate": 4.317215315146238e-05,
      "loss": 0.0232,
      "step": 1930
    },
    {
      "epoch": 2.961549470470375,
      "grad_norm": 0.12495076656341553,
      "learning_rate": 4.264452890239315e-05,
      "loss": 0.0223,
      "step": 1940
    },
    {
      "epoch": 2.9768151893903254,
      "grad_norm": 0.26308339834213257,
      "learning_rate": 4.211774053701952e-05,
      "loss": 0.0229,
      "step": 1950
    },
    {
      "epoch": 2.9768151893903254,
      "eval_loss": 0.030370745807886124,
      "eval_runtime": 179.8091,
      "eval_samples_per_second": 6.479,
      "eval_steps_per_second": 6.479,
      "step": 1950
    },
    {
      "epoch": 2.9920809083102755,
      "grad_norm": 0.21157263219356537,
      "learning_rate": 4.159184792014145e-05,
      "loss": 0.0163,
      "step": 1960
    },
    {
      "epoch": 3.007346627230226,
      "grad_norm": 0.062321823090314865,
      "learning_rate": 4.1066910814765016e-05,
      "loss": 0.0113,
      "step": 1970
    },
    {
      "epoch": 3.0226123461501766,
      "grad_norm": 0.1957968771457672,
      "learning_rate": 4.0542988875310995e-05,
      "loss": 0.0138,
      "step": 1980
    },
    {
      "epoch": 3.0378780650701267,
      "grad_norm": 0.12392223626375198,
      "learning_rate": 4.002014164083552e-05,
      "loss": 0.012,
      "step": 1990
    },
    {
      "epoch": 3.0531437839900772,
      "grad_norm": 0.09297648817300797,
      "learning_rate": 3.9498428528264204e-05,
      "loss": 0.0087,
      "step": 2000
    },
    {
      "epoch": 3.0531437839900772,
      "eval_loss": 0.032914210110902786,
      "eval_runtime": 179.7483,
      "eval_samples_per_second": 6.481,
      "eval_steps_per_second": 6.481,
      "step": 2000
    },
    {
      "epoch": 3.068409502910028,
      "grad_norm": 0.1610821634531021,
      "learning_rate": 3.89779088256397e-05,
      "loss": 0.0093,
      "step": 2010
    },
    {
      "epoch": 3.083675221829978,
      "grad_norm": 0.18227307498455048,
      "learning_rate": 3.845864168538437e-05,
      "loss": 0.0129,
      "step": 2020
    },
    {
      "epoch": 3.0989409407499284,
      "grad_norm": 0.26220542192459106,
      "learning_rate": 3.794068611757794e-05,
      "loss": 0.0136,
      "step": 2030
    },
    {
      "epoch": 3.114206659669879,
      "grad_norm": 0.15985815227031708,
      "learning_rate": 3.7424100983251695e-05,
      "loss": 0.0122,
      "step": 2040
    },
    {
      "epoch": 3.129472378589829,
      "grad_norm": 0.06649836152791977,
      "learning_rate": 3.6908944987699345e-05,
      "loss": 0.0073,
      "step": 2050
    },
    {
      "epoch": 3.129472378589829,
      "eval_loss": 0.033551715314388275,
      "eval_runtime": 179.8686,
      "eval_samples_per_second": 6.477,
      "eval_steps_per_second": 6.477,
      "step": 2050
    },
    {
      "epoch": 3.1447380975097796,
      "grad_norm": 0.11365962028503418,
      "learning_rate": 3.639527667380571e-05,
      "loss": 0.0084,
      "step": 2060
    },
    {
      "epoch": 3.16000381642973,
      "grad_norm": 0.1839054673910141,
      "learning_rate": 3.5883154415393885e-05,
      "loss": 0.0101,
      "step": 2070
    },
    {
      "epoch": 3.1752695353496803,
      "grad_norm": 0.20857231318950653,
      "learning_rate": 3.537263641059152e-05,
      "loss": 0.0107,
      "step": 2080
    },
    {
      "epoch": 3.190535254269631,
      "grad_norm": 0.15886227786540985,
      "learning_rate": 3.486378067521718e-05,
      "loss": 0.0114,
      "step": 2090
    },
    {
      "epoch": 3.2058009731895813,
      "grad_norm": 0.10763030499219894,
      "learning_rate": 3.435664503618732e-05,
      "loss": 0.0068,
      "step": 2100
    },
    {
      "epoch": 3.2058009731895813,
      "eval_loss": 0.03458699211478233,
      "eval_runtime": 179.8039,
      "eval_samples_per_second": 6.479,
      "eval_steps_per_second": 6.479,
      "step": 2100
    },
    {
      "epoch": 3.2210666921095314,
      "grad_norm": 0.15906715393066406,
      "learning_rate": 3.3851287124944756e-05,
      "loss": 0.0091,
      "step": 2110
    },
    {
      "epoch": 3.236332411029482,
      "grad_norm": 0.07679519057273865,
      "learning_rate": 3.334776437090944e-05,
      "loss": 0.0094,
      "step": 2120
    },
    {
      "epoch": 3.251598129949432,
      "grad_norm": 0.07010853290557861,
      "learning_rate": 3.2846133994952046e-05,
      "loss": 0.0091,
      "step": 2130
    },
    {
      "epoch": 3.2668638488693826,
      "grad_norm": 0.28941255807876587,
      "learning_rate": 3.234645300289136e-05,
      "loss": 0.0051,
      "step": 2140
    },
    {
      "epoch": 3.282129567789333,
      "grad_norm": 0.2712288796901703,
      "learning_rate": 3.1848778179016075e-05,
      "loss": 0.0099,
      "step": 2150
    },
    {
      "epoch": 3.282129567789333,
      "eval_loss": 0.03673892095685005,
      "eval_runtime": 179.9183,
      "eval_samples_per_second": 6.475,
      "eval_steps_per_second": 6.475,
      "step": 2150
    },
    {
      "epoch": 3.2973952867092833,
      "grad_norm": 0.3802885413169861,
      "learning_rate": 3.135316607963176e-05,
      "loss": 0.0108,
      "step": 2160
    },
    {
      "epoch": 3.312661005629234,
      "grad_norm": 0.19786600768566132,
      "learning_rate": 3.085967302663375e-05,
      "loss": 0.0082,
      "step": 2170
    },
    {
      "epoch": 3.3279267245491844,
      "grad_norm": 0.25580260157585144,
      "learning_rate": 3.0368355101106615e-05,
      "loss": 0.0153,
      "step": 2180
    },
    {
      "epoch": 3.3431924434691345,
      "grad_norm": 0.18568645417690277,
      "learning_rate": 2.987926813695116e-05,
      "loss": 0.0092,
      "step": 2190
    },
    {
      "epoch": 3.358458162389085,
      "grad_norm": 0.15619678795337677,
      "learning_rate": 2.939246771453924e-05,
      "loss": 0.0124,
      "step": 2200
    },
    {
      "epoch": 3.358458162389085,
      "eval_loss": 0.035190921276807785,
      "eval_runtime": 180.011,
      "eval_samples_per_second": 6.472,
      "eval_steps_per_second": 6.472,
      "step": 2200
    },
    {
      "epoch": 3.3737238813090356,
      "grad_norm": 0.2756540775299072,
      "learning_rate": 2.890800915439772e-05,
      "loss": 0.0112,
      "step": 2210
    },
    {
      "epoch": 3.3889896002289857,
      "grad_norm": 0.15437598526477814,
      "learning_rate": 2.842594751092159e-05,
      "loss": 0.0095,
      "step": 2220
    },
    {
      "epoch": 3.404255319148936,
      "grad_norm": 0.21479307115077972,
      "learning_rate": 2.794633756611776e-05,
      "loss": 0.0122,
      "step": 2230
    },
    {
      "epoch": 3.4195210380688867,
      "grad_norm": 0.21604545414447784,
      "learning_rate": 2.7469233823379347e-05,
      "loss": 0.0089,
      "step": 2240
    },
    {
      "epoch": 3.434786756988837,
      "grad_norm": 0.38754087686538696,
      "learning_rate": 2.6994690501292032e-05,
      "loss": 0.0087,
      "step": 2250
    },
    {
      "epoch": 3.434786756988837,
      "eval_loss": 0.03597886487841606,
      "eval_runtime": 179.9469,
      "eval_samples_per_second": 6.474,
      "eval_steps_per_second": 6.474,
      "step": 2250
    },
    {
      "epoch": 3.4500524759087874,
      "grad_norm": 0.2704348862171173,
      "learning_rate": 2.652276152747246e-05,
      "loss": 0.0072,
      "step": 2260
    },
    {
      "epoch": 3.465318194828738,
      "grad_norm": 0.27907001972198486,
      "learning_rate": 2.6053500532439968e-05,
      "loss": 0.0078,
      "step": 2270
    },
    {
      "epoch": 3.480583913748688,
      "grad_norm": 0.024898068979382515,
      "learning_rate": 2.5586960843521824e-05,
      "loss": 0.0098,
      "step": 2280
    },
    {
      "epoch": 3.4958496326686386,
      "grad_norm": 0.4165664315223694,
      "learning_rate": 2.5123195478793217e-05,
      "loss": 0.0106,
      "step": 2290
    },
    {
      "epoch": 3.5111153515885887,
      "grad_norm": 0.17015843093395233,
      "learning_rate": 2.466225714105202e-05,
      "loss": 0.0131,
      "step": 2300
    },
    {
      "epoch": 3.5111153515885887,
      "eval_loss": 0.03541536629199982,
      "eval_runtime": 179.5955,
      "eval_samples_per_second": 6.487,
      "eval_steps_per_second": 6.487,
      "step": 2300
    },
    {
      "epoch": 3.526381070508539,
      "grad_norm": 0.10995722562074661,
      "learning_rate": 2.420419821182982e-05,
      "loss": 0.0118,
      "step": 2310
    },
    {
      "epoch": 3.5416467894284898,
      "grad_norm": 0.1612035483121872,
      "learning_rate": 2.3749070745438996e-05,
      "loss": 0.0063,
      "step": 2320
    },
    {
      "epoch": 3.5569125083484403,
      "grad_norm": 0.15208382904529572,
      "learning_rate": 2.3296926463057396e-05,
      "loss": 0.0102,
      "step": 2330
    },
    {
      "epoch": 3.5721782272683904,
      "grad_norm": 0.19346974790096283,
      "learning_rate": 2.284781674685058e-05,
      "loss": 0.0079,
      "step": 2340
    },
    {
      "epoch": 3.587443946188341,
      "grad_norm": 0.43242356181144714,
      "learning_rate": 2.2401792634132708e-05,
      "loss": 0.0143,
      "step": 2350
    },
    {
      "epoch": 3.587443946188341,
      "eval_loss": 0.036229364573955536,
      "eval_runtime": 178.755,
      "eval_samples_per_second": 6.517,
      "eval_steps_per_second": 6.517,
      "step": 2350
    },
    {
      "epoch": 3.602709665108291,
      "grad_norm": 0.35156679153442383,
      "learning_rate": 2.195890481156666e-05,
      "loss": 0.0097,
      "step": 2360
    },
    {
      "epoch": 3.6179753840282416,
      "grad_norm": 0.09875629097223282,
      "learning_rate": 2.151920360940387e-05,
      "loss": 0.0112,
      "step": 2370
    },
    {
      "epoch": 3.633241102948192,
      "grad_norm": 0.26106587052345276,
      "learning_rate": 2.1082738995764785e-05,
      "loss": 0.0078,
      "step": 2380
    },
    {
      "epoch": 3.6485068218681422,
      "grad_norm": 0.15816567838191986,
      "learning_rate": 2.0649560570960465e-05,
      "loss": 0.0046,
      "step": 2390
    },
    {
      "epoch": 3.663772540788093,
      "grad_norm": 0.2872539460659027,
      "learning_rate": 2.0219717561855855e-05,
      "loss": 0.0066,
      "step": 2400
    },
    {
      "epoch": 3.663772540788093,
      "eval_loss": 0.036142896860837936,
      "eval_runtime": 177.8777,
      "eval_samples_per_second": 6.549,
      "eval_steps_per_second": 6.549,
      "step": 2400
    },
    {
      "epoch": 3.679038259708043,
      "grad_norm": 0.16133087873458862,
      "learning_rate": 1.9793258816275728e-05,
      "loss": 0.0112,
      "step": 2410
    },
    {
      "epoch": 3.6943039786279934,
      "grad_norm": 0.32570400834083557,
      "learning_rate": 1.9370232797453402e-05,
      "loss": 0.009,
      "step": 2420
    },
    {
      "epoch": 3.709569697547944,
      "grad_norm": 0.21927714347839355,
      "learning_rate": 1.8950687578523502e-05,
      "loss": 0.0085,
      "step": 2430
    },
    {
      "epoch": 3.7248354164678945,
      "grad_norm": 0.2975069284439087,
      "learning_rate": 1.853467083705869e-05,
      "loss": 0.0057,
      "step": 2440
    },
    {
      "epoch": 3.7401011353878446,
      "grad_norm": 0.19914095103740692,
      "learning_rate": 1.8122229849651716e-05,
      "loss": 0.0093,
      "step": 2450
    },
    {
      "epoch": 3.7401011353878446,
      "eval_loss": 0.03722745552659035,
      "eval_runtime": 177.626,
      "eval_samples_per_second": 6.559,
      "eval_steps_per_second": 6.559,
      "step": 2450
    },
    {
      "epoch": 3.755366854307795,
      "grad_norm": 0.20476055145263672,
      "learning_rate": 1.7713411486542707e-05,
      "loss": 0.0091,
      "step": 2460
    },
    {
      "epoch": 3.7706325732277453,
      "grad_norm": 0.22789469361305237,
      "learning_rate": 1.7308262206292897e-05,
      "loss": 0.0111,
      "step": 2470
    },
    {
      "epoch": 3.785898292147696,
      "grad_norm": 0.16632677614688873,
      "learning_rate": 1.6906828050504907e-05,
      "loss": 0.014,
      "step": 2480
    },
    {
      "epoch": 3.8011640110676463,
      "grad_norm": 0.23071546852588654,
      "learning_rate": 1.650915463859068e-05,
      "loss": 0.0092,
      "step": 2490
    },
    {
      "epoch": 3.8164297299875964,
      "grad_norm": 0.08320309221744537,
      "learning_rate": 1.6115287162587055e-05,
      "loss": 0.0074,
      "step": 2500
    },
    {
      "epoch": 3.8164297299875964,
      "eval_loss": 0.03621383011341095,
      "eval_runtime": 177.3254,
      "eval_samples_per_second": 6.57,
      "eval_steps_per_second": 6.57,
      "step": 2500
    },
    {
      "epoch": 3.831695448907547,
      "grad_norm": 0.07662249356508255,
      "learning_rate": 1.57252703820203e-05,
      "loss": 0.0069,
      "step": 2510
    },
    {
      "epoch": 3.8469611678274975,
      "grad_norm": 0.35770368576049805,
      "learning_rate": 1.5339148618819393e-05,
      "loss": 0.0093,
      "step": 2520
    },
    {
      "epoch": 3.8622268867474476,
      "grad_norm": 0.17595680058002472,
      "learning_rate": 1.4956965752279395e-05,
      "loss": 0.0105,
      "step": 2530
    },
    {
      "epoch": 3.877492605667398,
      "grad_norm": 0.28761520981788635,
      "learning_rate": 1.457876521407484e-05,
      "loss": 0.0124,
      "step": 2540
    },
    {
      "epoch": 3.8927583245873487,
      "grad_norm": 0.24026645720005035,
      "learning_rate": 1.4204589983324173e-05,
      "loss": 0.0081,
      "step": 2550
    },
    {
      "epoch": 3.8927583245873487,
      "eval_loss": 0.03659246861934662,
      "eval_runtime": 177.2547,
      "eval_samples_per_second": 6.572,
      "eval_steps_per_second": 6.572,
      "step": 2550
    },
    {
      "epoch": 3.908024043507299,
      "grad_norm": 0.1325196474790573,
      "learning_rate": 1.383448258170557e-05,
      "loss": 0.0097,
      "step": 2560
    },
    {
      "epoch": 3.9232897624272494,
      "grad_norm": 0.18720892071723938,
      "learning_rate": 1.3468485068624653e-05,
      "loss": 0.009,
      "step": 2570
    },
    {
      "epoch": 3.9385554813471995,
      "grad_norm": 0.03328974172472954,
      "learning_rate": 1.310663903643492e-05,
      "loss": 0.009,
      "step": 2580
    },
    {
      "epoch": 3.95382120026715,
      "grad_norm": 0.08329346030950546,
      "learning_rate": 1.2748985605711028e-05,
      "loss": 0.0087,
      "step": 2590
    },
    {
      "epoch": 3.9690869191871005,
      "grad_norm": 0.1863928586244583,
      "learning_rate": 1.2395565420575932e-05,
      "loss": 0.0088,
      "step": 2600
    },
    {
      "epoch": 3.9690869191871005,
      "eval_loss": 0.03702830523252487,
      "eval_runtime": 176.9895,
      "eval_samples_per_second": 6.582,
      "eval_steps_per_second": 6.582,
      "step": 2600
    },
    {
      "epoch": 3.984352638107051,
      "grad_norm": 0.1559509038925171,
      "learning_rate": 1.2046418644081903e-05,
      "loss": 0.0178,
      "step": 2610
    },
    {
      "epoch": 3.999618357027001,
      "grad_norm": 0.302651584148407,
      "learning_rate": 1.1701584953646505e-05,
      "loss": 0.0086,
      "step": 2620
    },
    {
      "epoch": 4.014884075946951,
      "grad_norm": 0.19416853785514832,
      "learning_rate": 1.1361103536543466e-05,
      "loss": 0.0047,
      "step": 2630
    },
    {
      "epoch": 4.030149794866902,
      "grad_norm": 0.0938921868801117,
      "learning_rate": 1.1025013085449527e-05,
      "loss": 0.0035,
      "step": 2640
    },
    {
      "epoch": 4.045415513786852,
      "grad_norm": 0.1689745932817459,
      "learning_rate": 1.0693351794047224e-05,
      "loss": 0.0052,
      "step": 2650
    },
    {
      "epoch": 4.045415513786852,
      "eval_loss": 0.037683501839637756,
      "eval_runtime": 176.8431,
      "eval_samples_per_second": 6.588,
      "eval_steps_per_second": 6.588,
      "step": 2650
    },
    {
      "epoch": 4.060681232706803,
      "grad_norm": 0.13139070570468903,
      "learning_rate": 1.036615735268468e-05,
      "loss": 0.0041,
      "step": 2660
    },
    {
      "epoch": 4.0759469516267535,
      "grad_norm": 0.04985181987285614,
      "learning_rate": 1.0043466944092272e-05,
      "loss": 0.0059,
      "step": 2670
    },
    {
      "epoch": 4.091212670546703,
      "grad_norm": 0.09271463006734848,
      "learning_rate": 9.72531723915726e-06,
      "loss": 0.0086,
      "step": 2680
    },
    {
      "epoch": 4.106478389466654,
      "grad_norm": 0.03634309768676758,
      "learning_rate": 9.411744392756405e-06,
      "loss": 0.0052,
      "step": 2690
    },
    {
      "epoch": 4.121744108386604,
      "grad_norm": 0.06921113282442093,
      "learning_rate": 9.102784039647339e-06,
      "loss": 0.0053,
      "step": 2700
    },
    {
      "epoch": 4.121744108386604,
      "eval_loss": 0.039026983082294464,
      "eval_runtime": 176.6518,
      "eval_samples_per_second": 6.595,
      "eval_steps_per_second": 6.595,
      "step": 2700
    },
    {
      "epoch": 4.137009827306555,
      "grad_norm": 0.046318791806697845,
      "learning_rate": 8.79847129041893e-06,
      "loss": 0.0039,
      "step": 2710
    },
    {
      "epoch": 4.152275546226505,
      "grad_norm": 0.07762674242258072,
      "learning_rate": 8.498840727501316e-06,
      "loss": 0.0049,
      "step": 2720
    },
    {
      "epoch": 4.167541265146456,
      "grad_norm": 0.0868455171585083,
      "learning_rate": 8.203926401235957e-06,
      "loss": 0.0043,
      "step": 2730
    },
    {
      "epoch": 4.1828069840664055,
      "grad_norm": 0.13933709263801575,
      "learning_rate": 7.913761826006017e-06,
      "loss": 0.0037,
      "step": 2740
    },
    {
      "epoch": 4.198072702986356,
      "grad_norm": 0.11513333767652512,
      "learning_rate": 7.628379976427868e-06,
      "loss": 0.0025,
      "step": 2750
    },
    {
      "epoch": 4.198072702986356,
      "eval_loss": 0.040155019611120224,
      "eval_runtime": 176.5921,
      "eval_samples_per_second": 6.597,
      "eval_steps_per_second": 6.597,
      "step": 2750
    },
    {
      "epoch": 4.213338421906307,
      "grad_norm": 0.3080286681652069,
      "learning_rate": 7.347813283603705e-06,
      "loss": 0.0067,
      "step": 2760
    },
    {
      "epoch": 4.228604140826257,
      "grad_norm": 0.06484497338533401,
      "learning_rate": 7.072093631436161e-06,
      "loss": 0.0041,
      "step": 2770
    },
    {
      "epoch": 4.243869859746208,
      "grad_norm": 0.023219596594572067,
      "learning_rate": 6.801252353004867e-06,
      "loss": 0.0039,
      "step": 2780
    },
    {
      "epoch": 4.259135578666158,
      "grad_norm": 0.13570910692214966,
      "learning_rate": 6.535320227005826e-06,
      "loss": 0.0059,
      "step": 2790
    },
    {
      "epoch": 4.274401297586108,
      "grad_norm": 0.0693252682685852,
      "learning_rate": 6.274327474253611e-06,
      "loss": 0.0027,
      "step": 2800
    },
    {
      "epoch": 4.274401297586108,
      "eval_loss": 0.04144105687737465,
      "eval_runtime": 176.6226,
      "eval_samples_per_second": 6.596,
      "eval_steps_per_second": 6.596,
      "step": 2800
    },
    {
      "epoch": 4.289667016506058,
      "grad_norm": 0.10616520047187805,
      "learning_rate": 6.018303754247112e-06,
      "loss": 0.0051,
      "step": 2810
    },
    {
      "epoch": 4.304932735426009,
      "grad_norm": 0.14562755823135376,
      "learning_rate": 5.767278161798911e-06,
      "loss": 0.0042,
      "step": 2820
    },
    {
      "epoch": 4.3201984543459595,
      "grad_norm": 0.1878853291273117,
      "learning_rate": 5.521279223729026e-06,
      "loss": 0.0035,
      "step": 2830
    },
    {
      "epoch": 4.33546417326591,
      "grad_norm": 0.17619742453098297,
      "learning_rate": 5.280334895622968e-06,
      "loss": 0.004,
      "step": 2840
    },
    {
      "epoch": 4.350729892185861,
      "grad_norm": 0.15715391933918,
      "learning_rate": 5.044472558654961e-06,
      "loss": 0.0076,
      "step": 2850
    },
    {
      "epoch": 4.350729892185861,
      "eval_loss": 0.04184148460626602,
      "eval_runtime": 176.611,
      "eval_samples_per_second": 6.596,
      "eval_steps_per_second": 6.596,
      "step": 2850
    },
    {
      "epoch": 4.36599561110581,
      "grad_norm": 0.05749718099832535,
      "learning_rate": 4.813719016476203e-06,
      "loss": 0.0029,
      "step": 2860
    },
    {
      "epoch": 4.381261330025761,
      "grad_norm": 0.27915164828300476,
      "learning_rate": 4.588100492168973e-06,
      "loss": 0.003,
      "step": 2870
    },
    {
      "epoch": 4.396527048945711,
      "grad_norm": 0.26044297218322754,
      "learning_rate": 4.367642625266511e-06,
      "loss": 0.0027,
      "step": 2880
    },
    {
      "epoch": 4.411792767865662,
      "grad_norm": 0.09349619597196579,
      "learning_rate": 4.1523704688394176e-06,
      "loss": 0.0042,
      "step": 2890
    },
    {
      "epoch": 4.427058486785612,
      "grad_norm": 0.06762149930000305,
      "learning_rate": 3.9423084866484884e-06,
      "loss": 0.0022,
      "step": 2900
    },
    {
      "epoch": 4.427058486785612,
      "eval_loss": 0.04281175881624222,
      "eval_runtime": 176.5277,
      "eval_samples_per_second": 6.6,
      "eval_steps_per_second": 6.6,
      "step": 2900
    },
    {
      "epoch": 4.442324205705562,
      "grad_norm": 0.15383104979991913,
      "learning_rate": 3.737480550364736e-06,
      "loss": 0.0062,
      "step": 2910
    },
    {
      "epoch": 4.457589924625513,
      "grad_norm": 0.21743904054164886,
      "learning_rate": 3.5379099368564817e-06,
      "loss": 0.0056,
      "step": 2920
    },
    {
      "epoch": 4.472855643545463,
      "grad_norm": 0.05257946625351906,
      "learning_rate": 3.3436193255442396e-06,
      "loss": 0.0034,
      "step": 2930
    },
    {
      "epoch": 4.488121362465414,
      "grad_norm": 0.05678296461701393,
      "learning_rate": 3.1546307958233214e-06,
      "loss": 0.0065,
      "step": 2940
    },
    {
      "epoch": 4.503387081385364,
      "grad_norm": 0.09029638022184372,
      "learning_rate": 2.9709658245547834e-06,
      "loss": 0.0036,
      "step": 2950
    },
    {
      "epoch": 4.503387081385364,
      "eval_loss": 0.04338888078927994,
      "eval_runtime": 176.4344,
      "eval_samples_per_second": 6.603,
      "eval_steps_per_second": 6.603,
      "step": 2950
    },
    {
      "epoch": 4.518652800305315,
      "grad_norm": 0.017809836193919182,
      "learning_rate": 2.792645283624712e-06,
      "loss": 0.0051,
      "step": 2960
    },
    {
      "epoch": 4.5339185192252645,
      "grad_norm": 0.08156479895114899,
      "learning_rate": 2.6196894375723645e-06,
      "loss": 0.0016,
      "step": 2970
    },
    {
      "epoch": 4.549184238145215,
      "grad_norm": 0.24031901359558105,
      "learning_rate": 2.452117941287246e-06,
      "loss": 0.0049,
      "step": 2980
    },
    {
      "epoch": 4.5644499570651655,
      "grad_norm": 0.2679056227207184,
      "learning_rate": 2.2899498377755566e-06,
      "loss": 0.0024,
      "step": 2990
    },
    {
      "epoch": 4.579715675985116,
      "grad_norm": 0.1409529745578766,
      "learning_rate": 2.1332035559960663e-06,
      "loss": 0.003,
      "step": 3000
    },
    {
      "epoch": 4.579715675985116,
      "eval_loss": 0.04322050139307976,
      "eval_runtime": 176.3389,
      "eval_samples_per_second": 6.607,
      "eval_steps_per_second": 6.607,
      "step": 3000
    },
    {
      "epoch": 4.594981394905067,
      "grad_norm": 0.3694174289703369,
      "learning_rate": 1.9818969087658735e-06,
      "loss": 0.0034,
      "step": 3010
    },
    {
      "epoch": 4.610247113825016,
      "grad_norm": 0.007240964099764824,
      "learning_rate": 1.8360470907361093e-06,
      "loss": 0.002,
      "step": 3020
    },
    {
      "epoch": 4.625512832744967,
      "grad_norm": 0.21716032922267914,
      "learning_rate": 1.6956706764379438e-06,
      "loss": 0.0046,
      "step": 3030
    },
    {
      "epoch": 4.640778551664917,
      "grad_norm": 0.10940048843622208,
      "learning_rate": 1.5607836183989921e-06,
      "loss": 0.0033,
      "step": 3040
    },
    {
      "epoch": 4.656044270584868,
      "grad_norm": 0.14919409155845642,
      "learning_rate": 1.4314012453305215e-06,
      "loss": 0.002,
      "step": 3050
    },
    {
      "epoch": 4.656044270584868,
      "eval_loss": 0.043372880667448044,
      "eval_runtime": 176.3259,
      "eval_samples_per_second": 6.607,
      "eval_steps_per_second": 6.607,
      "step": 3050
    },
    {
      "epoch": 4.6713099895048185,
      "grad_norm": 0.21336524188518524,
      "learning_rate": 1.3075382603854157e-06,
      "loss": 0.003,
      "step": 3060
    },
    {
      "epoch": 4.686575708424769,
      "grad_norm": 0.1394193321466446,
      "learning_rate": 1.1892087394873353e-06,
      "loss": 0.0034,
      "step": 3070
    },
    {
      "epoch": 4.701841427344719,
      "grad_norm": 0.02440819889307022,
      "learning_rate": 1.076426129731084e-06,
      "loss": 0.0036,
      "step": 3080
    },
    {
      "epoch": 4.717107146264669,
      "grad_norm": 0.14676684141159058,
      "learning_rate": 9.692032478545e-07,
      "loss": 0.0069,
      "step": 3090
    },
    {
      "epoch": 4.73237286518462,
      "grad_norm": 0.01789388246834278,
      "learning_rate": 8.675522787819023e-07,
      "loss": 0.0078,
      "step": 3100
    },
    {
      "epoch": 4.73237286518462,
      "eval_loss": 0.043460384011268616,
      "eval_runtime": 176.2874,
      "eval_samples_per_second": 6.609,
      "eval_steps_per_second": 6.609,
      "step": 3100
    },
    {
      "epoch": 4.74763858410457,
      "grad_norm": 0.04277348518371582,
      "learning_rate": 7.714847742394337e-07,
      "loss": 0.0053,
      "step": 3110
    },
    {
      "epoch": 4.762904303024521,
      "grad_norm": 0.26189810037612915,
      "learning_rate": 6.810116514422593e-07,
      "loss": 0.005,
      "step": 3120
    },
    {
      "epoch": 4.7781700219444705,
      "grad_norm": 0.054308198392391205,
      "learning_rate": 5.961431918539817e-07,
      "loss": 0.0032,
      "step": 3130
    },
    {
      "epoch": 4.793435740864421,
      "grad_norm": 0.10629548877477646,
      "learning_rate": 5.16889040018187e-07,
      "loss": 0.0055,
      "step": 3140
    },
    {
      "epoch": 4.808701459784372,
      "grad_norm": 0.05638623237609863,
      "learning_rate": 4.432582024624543e-07,
      "loss": 0.0033,
      "step": 3150
    },
    {
      "epoch": 4.808701459784372,
      "eval_loss": 0.04357059299945831,
      "eval_runtime": 176.3321,
      "eval_samples_per_second": 6.607,
      "eval_steps_per_second": 6.607,
      "step": 3150
    },
    {
      "epoch": 4.823967178704322,
      "grad_norm": 0.035809215158224106,
      "learning_rate": 3.7525904667484737e-07,
      "loss": 0.0029,
      "step": 3160
    },
    {
      "epoch": 4.839232897624273,
      "grad_norm": 0.25203967094421387,
      "learning_rate": 3.128993001530245e-07,
      "loss": 0.0059,
      "step": 3170
    },
    {
      "epoch": 4.854498616544223,
      "grad_norm": 0.04179414361715317,
      "learning_rate": 2.5618604952605816e-07,
      "loss": 0.0046,
      "step": 3180
    },
    {
      "epoch": 4.869764335464174,
      "grad_norm": 0.14175380766391754,
      "learning_rate": 2.0512573974912908e-07,
      "loss": 0.0038,
      "step": 3190
    },
    {
      "epoch": 4.885030054384123,
      "grad_norm": 0.20164638757705688,
      "learning_rate": 1.597241733711008e-07,
      "loss": 0.0055,
      "step": 3200
    },
    {
      "epoch": 4.885030054384123,
      "eval_loss": 0.043629422783851624,
      "eval_runtime": 176.2217,
      "eval_samples_per_second": 6.611,
      "eval_steps_per_second": 6.611,
      "step": 3200
    },
    {
      "epoch": 4.900295773304074,
      "grad_norm": 0.10687878727912903,
      "learning_rate": 1.1998650987510295e-07,
      "loss": 0.0035,
      "step": 3210
    },
    {
      "epoch": 4.9155614922240245,
      "grad_norm": 0.010824548080563545,
      "learning_rate": 8.591726509222242e-08,
      "loss": 0.0074,
      "step": 3220
    },
    {
      "epoch": 4.930827211143975,
      "grad_norm": 0.13975496590137482,
      "learning_rate": 5.752031068830266e-08,
      "loss": 0.0041,
      "step": 3230
    },
    {
      "epoch": 4.946092930063926,
      "grad_norm": 0.08203299343585968,
      "learning_rate": 3.4798873723984605e-08,
      "loss": 0.0032,
      "step": 3240
    },
    {
      "epoch": 4.961358648983875,
      "grad_norm": 0.22268716990947723,
      "learning_rate": 1.7755536287944464e-08,
      "loss": 0.006,
      "step": 3250
    },
    {
      "epoch": 4.961358648983875,
      "eval_loss": 0.043610408902168274,
      "eval_runtime": 176.2018,
      "eval_samples_per_second": 6.612,
      "eval_steps_per_second": 6.612,
      "step": 3250
    },
    {
      "epoch": 4.976624367903826,
      "grad_norm": 0.07115557044744492,
      "learning_rate": 6.3922352034895276e-09,
      "loss": 0.0103,
      "step": 3260
    },
    {
      "epoch": 4.991890086823776,
      "grad_norm": 0.18499644100666046,
      "learning_rate": 7.102618084620094e-10,
      "loss": 0.0049,
      "step": 3270
    },
    {
      "epoch": 4.999522946283752,
      "step": 3275,
      "total_flos": 8.8535579239868e+17,
      "train_loss": 0.048749475333526845,
      "train_runtime": 37525.7545,
      "train_samples_per_second": 1.397,
      "train_steps_per_second": 0.087
    }
  ],
  "logging_steps": 10,
  "max_steps": 3275,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.8535579239868e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}