{
  "best_metric": 0.03486299887299538,
  "best_model_checkpoint": "saves/psy-course/Llama3-OpenBioLLM-8B/train/fold4/checkpoint-1300",
  "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": 5.458767414093018,
      "learning_rate": 3.0487804878048782e-06,
      "loss": 1.4201,
      "step": 10
    },
    {
      "epoch": 0.030531437839900772,
      "grad_norm": 5.880963325500488,
      "learning_rate": 6.0975609756097564e-06,
      "loss": 1.3361,
      "step": 20
    },
    {
      "epoch": 0.04579715675985116,
      "grad_norm": 4.0639777183532715,
      "learning_rate": 9.146341463414634e-06,
      "loss": 1.1539,
      "step": 30
    },
    {
      "epoch": 0.061062875679801544,
      "grad_norm": 3.002682685852051,
      "learning_rate": 1.2195121951219513e-05,
      "loss": 0.8971,
      "step": 40
    },
    {
      "epoch": 0.07632859459975193,
      "grad_norm": 1.940627098083496,
      "learning_rate": 1.524390243902439e-05,
      "loss": 0.4799,
      "step": 50
    },
    {
      "epoch": 0.07632859459975193,
      "eval_loss": 0.2933095097541809,
      "eval_runtime": 159.4102,
      "eval_samples_per_second": 7.308,
      "eval_steps_per_second": 7.308,
      "step": 50
    },
    {
      "epoch": 0.09159431351970232,
      "grad_norm": 1.761772632598877,
      "learning_rate": 1.8292682926829268e-05,
      "loss": 0.2509,
      "step": 60
    },
    {
      "epoch": 0.10686003243965271,
      "grad_norm": 1.450770616531372,
      "learning_rate": 2.134146341463415e-05,
      "loss": 0.2104,
      "step": 70
    },
    {
      "epoch": 0.12212575135960309,
      "grad_norm": 1.4452862739562988,
      "learning_rate": 2.4390243902439026e-05,
      "loss": 0.1552,
      "step": 80
    },
    {
      "epoch": 0.13739147027955348,
      "grad_norm": 1.2762833833694458,
      "learning_rate": 2.7439024390243906e-05,
      "loss": 0.1008,
      "step": 90
    },
    {
      "epoch": 0.15265718919950386,
      "grad_norm": 0.8838859796524048,
      "learning_rate": 3.048780487804878e-05,
      "loss": 0.1053,
      "step": 100
    },
    {
      "epoch": 0.15265718919950386,
      "eval_loss": 0.07661888748407364,
      "eval_runtime": 159.8002,
      "eval_samples_per_second": 7.29,
      "eval_steps_per_second": 7.29,
      "step": 100
    },
    {
      "epoch": 0.16792290811945426,
      "grad_norm": 0.9274832606315613,
      "learning_rate": 3.353658536585366e-05,
      "loss": 0.0829,
      "step": 110
    },
    {
      "epoch": 0.18318862703940464,
      "grad_norm": 0.9330371618270874,
      "learning_rate": 3.6585365853658535e-05,
      "loss": 0.0803,
      "step": 120
    },
    {
      "epoch": 0.19845434595935502,
      "grad_norm": 1.1532729864120483,
      "learning_rate": 3.9634146341463416e-05,
      "loss": 0.0772,
      "step": 130
    },
    {
      "epoch": 0.21372006487930542,
      "grad_norm": 0.9599048495292664,
      "learning_rate": 4.26829268292683e-05,
      "loss": 0.0777,
      "step": 140
    },
    {
      "epoch": 0.2289857837992558,
      "grad_norm": 1.0932921171188354,
      "learning_rate": 4.573170731707318e-05,
      "loss": 0.0594,
      "step": 150
    },
    {
      "epoch": 0.2289857837992558,
      "eval_loss": 0.072237528860569,
      "eval_runtime": 159.5845,
      "eval_samples_per_second": 7.3,
      "eval_steps_per_second": 7.3,
      "step": 150
    },
    {
      "epoch": 0.24425150271920618,
      "grad_norm": 0.9019838571548462,
      "learning_rate": 4.878048780487805e-05,
      "loss": 0.0845,
      "step": 160
    },
    {
      "epoch": 0.2595172216391566,
      "grad_norm": 1.1281777620315552,
      "learning_rate": 5.182926829268293e-05,
      "loss": 0.0777,
      "step": 170
    },
    {
      "epoch": 0.27478294055910696,
      "grad_norm": 0.8680335283279419,
      "learning_rate": 5.487804878048781e-05,
      "loss": 0.0886,
      "step": 180
    },
    {
      "epoch": 0.29004865947905734,
      "grad_norm": 0.7780885696411133,
      "learning_rate": 5.792682926829268e-05,
      "loss": 0.0648,
      "step": 190
    },
    {
      "epoch": 0.3053143783990077,
      "grad_norm": 0.784793496131897,
      "learning_rate": 6.097560975609756e-05,
      "loss": 0.0586,
      "step": 200
    },
    {
      "epoch": 0.3053143783990077,
      "eval_loss": 0.053653329610824585,
      "eval_runtime": 159.5655,
      "eval_samples_per_second": 7.301,
      "eval_steps_per_second": 7.301,
      "step": 200
    },
    {
      "epoch": 0.3205800973189581,
      "grad_norm": 0.8396239876747131,
      "learning_rate": 6.402439024390244e-05,
      "loss": 0.0569,
      "step": 210
    },
    {
      "epoch": 0.3358458162389085,
      "grad_norm": 1.3665101528167725,
      "learning_rate": 6.707317073170732e-05,
      "loss": 0.0523,
      "step": 220
    },
    {
      "epoch": 0.3511115351588589,
      "grad_norm": 0.4989739954471588,
      "learning_rate": 7.012195121951219e-05,
      "loss": 0.045,
      "step": 230
    },
    {
      "epoch": 0.3663772540788093,
      "grad_norm": 0.5166968703269958,
      "learning_rate": 7.317073170731707e-05,
      "loss": 0.0667,
      "step": 240
    },
    {
      "epoch": 0.38164297299875966,
      "grad_norm": 0.7004739046096802,
      "learning_rate": 7.621951219512195e-05,
      "loss": 0.066,
      "step": 250
    },
    {
      "epoch": 0.38164297299875966,
      "eval_loss": 0.0550156868994236,
      "eval_runtime": 159.9016,
      "eval_samples_per_second": 7.286,
      "eval_steps_per_second": 7.286,
      "step": 250
    },
    {
      "epoch": 0.39690869191871003,
      "grad_norm": 1.4189491271972656,
      "learning_rate": 7.926829268292683e-05,
      "loss": 0.0635,
      "step": 260
    },
    {
      "epoch": 0.4121744108386604,
      "grad_norm": 0.7976926565170288,
      "learning_rate": 8.231707317073171e-05,
      "loss": 0.0653,
      "step": 270
    },
    {
      "epoch": 0.42744012975861084,
      "grad_norm": 0.3455027937889099,
      "learning_rate": 8.53658536585366e-05,
      "loss": 0.0568,
      "step": 280
    },
    {
      "epoch": 0.4427058486785612,
      "grad_norm": 0.564358115196228,
      "learning_rate": 8.841463414634147e-05,
      "loss": 0.0629,
      "step": 290
    },
    {
      "epoch": 0.4579715675985116,
      "grad_norm": 0.3834143877029419,
      "learning_rate": 9.146341463414635e-05,
      "loss": 0.0553,
      "step": 300
    },
    {
      "epoch": 0.4579715675985116,
      "eval_loss": 0.055096790194511414,
      "eval_runtime": 159.7991,
      "eval_samples_per_second": 7.29,
      "eval_steps_per_second": 7.29,
      "step": 300
    },
    {
      "epoch": 0.473237286518462,
      "grad_norm": 0.8542683720588684,
      "learning_rate": 9.451219512195122e-05,
      "loss": 0.067,
      "step": 310
    },
    {
      "epoch": 0.48850300543841235,
      "grad_norm": 0.41747626662254333,
      "learning_rate": 9.75609756097561e-05,
      "loss": 0.05,
      "step": 320
    },
    {
      "epoch": 0.5037687243583627,
      "grad_norm": 0.5380941033363342,
      "learning_rate": 9.999988635788465e-05,
      "loss": 0.0552,
      "step": 330
    },
    {
      "epoch": 0.5190344432783132,
      "grad_norm": 0.7595194578170776,
      "learning_rate": 9.999590893808788e-05,
      "loss": 0.0437,
      "step": 340
    },
    {
      "epoch": 0.5343001621982635,
      "grad_norm": 0.8340535759925842,
      "learning_rate": 9.998624992909386e-05,
      "loss": 0.0468,
      "step": 350
    },
    {
      "epoch": 0.5343001621982635,
      "eval_loss": 0.04763290658593178,
      "eval_runtime": 159.9408,
      "eval_samples_per_second": 7.284,
      "eval_steps_per_second": 7.284,
      "step": 350
    },
    {
      "epoch": 0.5495658811182139,
      "grad_norm": 0.32083097100257874,
      "learning_rate": 9.997091042856284e-05,
      "loss": 0.0503,
      "step": 360
    },
    {
      "epoch": 0.5648316000381643,
      "grad_norm": 0.26985791325569153,
      "learning_rate": 9.994989217969224e-05,
      "loss": 0.0576,
      "step": 370
    },
    {
      "epoch": 0.5800973189581147,
      "grad_norm": 1.5231789350509644,
      "learning_rate": 9.992319757101863e-05,
      "loss": 0.0706,
      "step": 380
    },
    {
      "epoch": 0.5953630378780651,
      "grad_norm": 0.46429556608200073,
      "learning_rate": 9.98908296361462e-05,
      "loss": 0.0597,
      "step": 390
    },
    {
      "epoch": 0.6106287567980154,
      "grad_norm": 0.5127663016319275,
      "learning_rate": 9.98527920534021e-05,
      "loss": 0.0515,
      "step": 400
    },
    {
      "epoch": 0.6106287567980154,
      "eval_loss": 0.04808073118329048,
      "eval_runtime": 159.9539,
      "eval_samples_per_second": 7.283,
      "eval_steps_per_second": 7.283,
      "step": 400
    },
    {
      "epoch": 0.6258944757179659,
      "grad_norm": 0.6281822323799133,
      "learning_rate": 9.980908914541844e-05,
      "loss": 0.0517,
      "step": 410
    },
    {
      "epoch": 0.6411601946379162,
      "grad_norm": 0.4508628249168396,
      "learning_rate": 9.975972587864095e-05,
      "loss": 0.0514,
      "step": 420
    },
    {
      "epoch": 0.6564259135578666,
      "grad_norm": 0.8544013500213623,
      "learning_rate": 9.970470786276467e-05,
      "loss": 0.041,
      "step": 430
    },
    {
      "epoch": 0.671691632477817,
      "grad_norm": 0.3640694320201874,
      "learning_rate": 9.964404135009648e-05,
      "loss": 0.0507,
      "step": 440
    },
    {
      "epoch": 0.6869573513977674,
      "grad_norm": 0.203624427318573,
      "learning_rate": 9.957773323484454e-05,
      "loss": 0.0562,
      "step": 450
    },
    {
      "epoch": 0.6869573513977674,
      "eval_loss": 0.04656098783016205,
      "eval_runtime": 159.9758,
      "eval_samples_per_second": 7.282,
      "eval_steps_per_second": 7.282,
      "step": 450
    },
    {
      "epoch": 0.7022230703177178,
      "grad_norm": 0.3792983591556549,
      "learning_rate": 9.950579105233483e-05,
      "loss": 0.0496,
      "step": 460
    },
    {
      "epoch": 0.7174887892376681,
      "grad_norm": 0.3796481490135193,
      "learning_rate": 9.94282229781548e-05,
      "loss": 0.0425,
      "step": 470
    },
    {
      "epoch": 0.7327545081576186,
      "grad_norm": 0.7942231297492981,
      "learning_rate": 9.934503782722438e-05,
      "loss": 0.0541,
      "step": 480
    },
    {
      "epoch": 0.748020227077569,
      "grad_norm": 0.2960570156574249,
      "learning_rate": 9.925624505279411e-05,
      "loss": 0.0369,
      "step": 490
    },
    {
      "epoch": 0.7632859459975193,
      "grad_norm": 0.2349403351545334,
      "learning_rate": 9.916185474537098e-05,
      "loss": 0.0347,
      "step": 500
    },
    {
      "epoch": 0.7632859459975193,
      "eval_loss": 0.04013669118285179,
      "eval_runtime": 159.8253,
      "eval_samples_per_second": 7.289,
      "eval_steps_per_second": 7.289,
      "step": 500
    },
    {
      "epoch": 0.7785516649174697,
      "grad_norm": 0.6597937345504761,
      "learning_rate": 9.906187763157168e-05,
      "loss": 0.0491,
      "step": 510
    },
    {
      "epoch": 0.7938173838374201,
      "grad_norm": 0.38688021898269653,
      "learning_rate": 9.895632507290362e-05,
      "loss": 0.0526,
      "step": 520
    },
    {
      "epoch": 0.8090831027573705,
      "grad_norm": 0.3897739052772522,
      "learning_rate": 9.884520906447379e-05,
      "loss": 0.0397,
      "step": 530
    },
    {
      "epoch": 0.8243488216773208,
      "grad_norm": 0.5511626601219177,
      "learning_rate": 9.872854223362562e-05,
      "loss": 0.0539,
      "step": 540
    },
    {
      "epoch": 0.8396145405972713,
      "grad_norm": 0.5939027667045593,
      "learning_rate": 9.860633783850406e-05,
      "loss": 0.0417,
      "step": 550
    },
    {
      "epoch": 0.8396145405972713,
      "eval_loss": 0.04075213149189949,
      "eval_runtime": 159.8385,
      "eval_samples_per_second": 7.289,
      "eval_steps_per_second": 7.289,
      "step": 550
    },
    {
      "epoch": 0.8548802595172217,
      "grad_norm": 0.25341716408729553,
      "learning_rate": 9.847860976654879e-05,
      "loss": 0.0475,
      "step": 560
    },
    {
      "epoch": 0.870145978437172,
      "grad_norm": 0.2727987468242645,
      "learning_rate": 9.834537253291616e-05,
      "loss": 0.0445,
      "step": 570
    },
    {
      "epoch": 0.8854116973571224,
      "grad_norm": 1.1958130598068237,
      "learning_rate": 9.820664127882957e-05,
      "loss": 0.0583,
      "step": 580
    },
    {
      "epoch": 0.9006774162770728,
      "grad_norm": 0.5200381875038147,
      "learning_rate": 9.806243176985888e-05,
      "loss": 0.0517,
      "step": 590
    },
    {
      "epoch": 0.9159431351970232,
      "grad_norm": 0.4853935241699219,
      "learning_rate": 9.791276039412875e-05,
      "loss": 0.0466,
      "step": 600
    },
    {
      "epoch": 0.9159431351970232,
      "eval_loss": 0.03929273039102554,
      "eval_runtime": 159.8836,
      "eval_samples_per_second": 7.287,
      "eval_steps_per_second": 7.287,
      "step": 600
    },
    {
      "epoch": 0.9312088541169735,
      "grad_norm": 0.5579703450202942,
      "learning_rate": 9.775764416045628e-05,
      "loss": 0.0625,
      "step": 610
    },
    {
      "epoch": 0.946474573036924,
      "grad_norm": 0.3649356961250305,
      "learning_rate": 9.759710069641814e-05,
      "loss": 0.0485,
      "step": 620
    },
    {
      "epoch": 0.9617402919568744,
      "grad_norm": 0.3876245319843292,
      "learning_rate": 9.743114824634734e-05,
      "loss": 0.0321,
      "step": 630
    },
    {
      "epoch": 0.9770060108768247,
      "grad_norm": 0.4524376392364502,
      "learning_rate": 9.725980566925989e-05,
      "loss": 0.0453,
      "step": 640
    },
    {
      "epoch": 0.9922717297967751,
      "grad_norm": 0.27509111166000366,
      "learning_rate": 9.708309243671165e-05,
      "loss": 0.0346,
      "step": 650
    },
    {
      "epoch": 0.9922717297967751,
      "eval_loss": 0.040449682623147964,
      "eval_runtime": 159.9169,
      "eval_samples_per_second": 7.285,
      "eval_steps_per_second": 7.285,
      "step": 650
    },
    {
      "epoch": 1.0075374487167255,
      "grad_norm": 0.21305875480175018,
      "learning_rate": 9.690102863058563e-05,
      "loss": 0.0335,
      "step": 660
    },
    {
      "epoch": 1.0228031676366758,
      "grad_norm": 0.3463568091392517,
      "learning_rate": 9.67136349408098e-05,
      "loss": 0.0349,
      "step": 670
    },
    {
      "epoch": 1.0380688865566263,
      "grad_norm": 0.25874170660972595,
      "learning_rate": 9.652093266300583e-05,
      "loss": 0.0403,
      "step": 680
    },
    {
      "epoch": 1.0533346054765766,
      "grad_norm": 0.1967257261276245,
      "learning_rate": 9.632294369606916e-05,
      "loss": 0.0254,
      "step": 690
    },
    {
      "epoch": 1.068600324396527,
      "grad_norm": 0.3825061619281769,
      "learning_rate": 9.61196905396802e-05,
      "loss": 0.035,
      "step": 700
    },
    {
      "epoch": 1.068600324396527,
      "eval_loss": 0.03885221853852272,
      "eval_runtime": 159.557,
      "eval_samples_per_second": 7.301,
      "eval_steps_per_second": 7.301,
      "step": 700
    },
    {
      "epoch": 1.0838660433164775,
      "grad_norm": 0.32184284925460815,
      "learning_rate": 9.591119629174764e-05,
      "loss": 0.0431,
      "step": 710
    },
    {
      "epoch": 1.0991317622364278,
      "grad_norm": 0.2665984332561493,
      "learning_rate": 9.569748464578343e-05,
      "loss": 0.0271,
      "step": 720
    },
    {
      "epoch": 1.1143974811563782,
      "grad_norm": 0.35844355821609497,
      "learning_rate": 9.54785798882103e-05,
      "loss": 0.029,
      "step": 730
    },
    {
      "epoch": 1.1296632000763287,
      "grad_norm": 0.2451615184545517,
      "learning_rate": 9.525450689560181e-05,
      "loss": 0.0301,
      "step": 740
    },
    {
      "epoch": 1.144928918996279,
      "grad_norm": 0.29228684306144714,
      "learning_rate": 9.502529113185532e-05,
      "loss": 0.0408,
      "step": 750
    },
    {
      "epoch": 1.144928918996279,
      "eval_loss": 0.038966357707977295,
      "eval_runtime": 158.592,
      "eval_samples_per_second": 7.346,
      "eval_steps_per_second": 7.346,
      "step": 750
    },
    {
      "epoch": 1.1601946379162293,
      "grad_norm": 0.25255903601646423,
      "learning_rate": 9.479095864529828e-05,
      "loss": 0.0323,
      "step": 760
    },
    {
      "epoch": 1.1754603568361797,
      "grad_norm": 0.245975524187088,
      "learning_rate": 9.455153606572806e-05,
      "loss": 0.0296,
      "step": 770
    },
    {
      "epoch": 1.1907260757561302,
      "grad_norm": 0.271707683801651,
      "learning_rate": 9.430705060138569e-05,
      "loss": 0.0337,
      "step": 780
    },
    {
      "epoch": 1.2059917946760805,
      "grad_norm": 0.6799024939537048,
      "learning_rate": 9.405753003586395e-05,
      "loss": 0.0281,
      "step": 790
    },
    {
      "epoch": 1.2212575135960309,
      "grad_norm": 0.2844121754169464,
      "learning_rate": 9.38030027249499e-05,
      "loss": 0.0324,
      "step": 800
    },
    {
      "epoch": 1.2212575135960309,
      "eval_loss": 0.04265734553337097,
      "eval_runtime": 157.6777,
      "eval_samples_per_second": 7.388,
      "eval_steps_per_second": 7.388,
      "step": 800
    },
    {
      "epoch": 1.2365232325159814,
      "grad_norm": 0.3621199131011963,
      "learning_rate": 9.354349759340263e-05,
      "loss": 0.0382,
      "step": 810
    },
    {
      "epoch": 1.2517889514359317,
      "grad_norm": 0.4844062030315399,
      "learning_rate": 9.327904413166615e-05,
      "loss": 0.0353,
      "step": 820
    },
    {
      "epoch": 1.267054670355882,
      "grad_norm": 0.4185052216053009,
      "learning_rate": 9.300967239251798e-05,
      "loss": 0.0352,
      "step": 830
    },
    {
      "epoch": 1.2823203892758324,
      "grad_norm": 0.21575963497161865,
      "learning_rate": 9.27354129876541e-05,
      "loss": 0.0344,
      "step": 840
    },
    {
      "epoch": 1.297586108195783,
      "grad_norm": 0.18729855120182037,
      "learning_rate": 9.245629708421008e-05,
      "loss": 0.0331,
      "step": 850
    },
    {
      "epoch": 1.297586108195783,
      "eval_loss": 0.038167908787727356,
      "eval_runtime": 157.202,
      "eval_samples_per_second": 7.411,
      "eval_steps_per_second": 7.411,
      "step": 850
    },
    {
      "epoch": 1.3128518271157332,
      "grad_norm": 0.505947470664978,
      "learning_rate": 9.217235640121926e-05,
      "loss": 0.0336,
      "step": 860
    },
    {
      "epoch": 1.3281175460356835,
      "grad_norm": 0.6576479077339172,
      "learning_rate": 9.188362320600812e-05,
      "loss": 0.0418,
      "step": 870
    },
    {
      "epoch": 1.343383264955634,
      "grad_norm": 0.45256391167640686,
      "learning_rate": 9.159013031052943e-05,
      "loss": 0.0349,
      "step": 880
    },
    {
      "epoch": 1.3586489838755844,
      "grad_norm": 0.4001959264278412,
      "learning_rate": 9.129191106763346e-05,
      "loss": 0.03,
      "step": 890
    },
    {
      "epoch": 1.3739147027955347,
      "grad_norm": 0.2914159297943115,
      "learning_rate": 9.098899936727771e-05,
      "loss": 0.0335,
      "step": 900
    },
    {
      "epoch": 1.3739147027955347,
      "eval_loss": 0.04329466447234154,
      "eval_runtime": 156.5773,
      "eval_samples_per_second": 7.44,
      "eval_steps_per_second": 7.44,
      "step": 900
    },
    {
      "epoch": 1.3891804217154853,
      "grad_norm": 0.26559117436408997,
      "learning_rate": 9.068142963267558e-05,
      "loss": 0.0246,
      "step": 910
    },
    {
      "epoch": 1.4044461406354356,
      "grad_norm": 0.3729347586631775,
      "learning_rate": 9.036923681638463e-05,
      "loss": 0.0331,
      "step": 920
    },
    {
      "epoch": 1.419711859555386,
      "grad_norm": 0.2612379491329193,
      "learning_rate": 9.00524563963343e-05,
      "loss": 0.0242,
      "step": 930
    },
    {
      "epoch": 1.4349775784753362,
      "grad_norm": 0.4376416802406311,
      "learning_rate": 8.973112437179436e-05,
      "loss": 0.0375,
      "step": 940
    },
    {
      "epoch": 1.4502432973952868,
      "grad_norm": 0.2591637074947357,
      "learning_rate": 8.940527725928383e-05,
      "loss": 0.034,
      "step": 950
    },
    {
      "epoch": 1.4502432973952868,
      "eval_loss": 0.037408314645290375,
      "eval_runtime": 156.3423,
      "eval_samples_per_second": 7.452,
      "eval_steps_per_second": 7.452,
      "step": 950
    },
    {
      "epoch": 1.4655090163152371,
      "grad_norm": 0.3151344060897827,
      "learning_rate": 8.90749520884212e-05,
      "loss": 0.0311,
      "step": 960
    },
    {
      "epoch": 1.4807747352351874,
      "grad_norm": 0.30333495140075684,
      "learning_rate": 8.874018639771637e-05,
      "loss": 0.0323,
      "step": 970
    },
    {
      "epoch": 1.4960404541551378,
      "grad_norm": 0.318057119846344,
      "learning_rate": 8.840101823030471e-05,
      "loss": 0.0293,
      "step": 980
    },
    {
      "epoch": 1.5113061730750883,
      "grad_norm": 0.4195692241191864,
      "learning_rate": 8.805748612962382e-05,
      "loss": 0.0396,
      "step": 990
    },
    {
      "epoch": 1.5265718919950386,
      "grad_norm": 0.1193198561668396,
      "learning_rate": 8.77096291350334e-05,
      "loss": 0.0253,
      "step": 1000
    },
    {
      "epoch": 1.5265718919950386,
      "eval_loss": 0.03985153138637543,
      "eval_runtime": 156.2778,
      "eval_samples_per_second": 7.455,
      "eval_steps_per_second": 7.455,
      "step": 1000
    },
    {
      "epoch": 1.541837610914989,
      "grad_norm": 0.2724255323410034,
      "learning_rate": 8.735748677737874e-05,
      "loss": 0.0356,
      "step": 1010
    },
    {
      "epoch": 1.5571033298349395,
      "grad_norm": 0.6145180463790894,
      "learning_rate": 8.700109907449845e-05,
      "loss": 0.0266,
      "step": 1020
    },
    {
      "epoch": 1.5723690487548898,
      "grad_norm": 0.26061272621154785,
      "learning_rate": 8.66405065266768e-05,
      "loss": 0.0277,
      "step": 1030
    },
    {
      "epoch": 1.5876347676748401,
      "grad_norm": 0.22841082513332367,
      "learning_rate": 8.627575011204115e-05,
      "loss": 0.0293,
      "step": 1040
    },
    {
      "epoch": 1.6029004865947907,
      "grad_norm": 0.3250572681427002,
      "learning_rate": 8.590687128190516e-05,
      "loss": 0.0299,
      "step": 1050
    },
    {
      "epoch": 1.6029004865947907,
      "eval_loss": 0.03824828937649727,
      "eval_runtime": 156.1357,
      "eval_samples_per_second": 7.461,
      "eval_steps_per_second": 7.461,
      "step": 1050
    },
    {
      "epoch": 1.618166205514741,
      "grad_norm": 0.09818235039710999,
      "learning_rate": 8.553391195605833e-05,
      "loss": 0.0271,
      "step": 1060
    },
    {
      "epoch": 1.6334319244346913,
      "grad_norm": 0.46138861775398254,
      "learning_rate": 8.515691451800205e-05,
      "loss": 0.04,
      "step": 1070
    },
    {
      "epoch": 1.6486976433546419,
      "grad_norm": 0.34611156582832336,
      "learning_rate": 8.477592181013316e-05,
      "loss": 0.0268,
      "step": 1080
    },
    {
      "epoch": 1.663963362274592,
      "grad_norm": 0.3302578032016754,
      "learning_rate": 8.439097712887531e-05,
      "loss": 0.03,
      "step": 1090
    },
    {
      "epoch": 1.6792290811945425,
      "grad_norm": 0.395153284072876,
      "learning_rate": 8.400212421975865e-05,
      "loss": 0.0534,
      "step": 1100
    },
    {
      "epoch": 1.6792290811945425,
      "eval_loss": 0.04240264743566513,
      "eval_runtime": 156.1164,
      "eval_samples_per_second": 7.462,
      "eval_steps_per_second": 7.462,
      "step": 1100
    },
    {
      "epoch": 1.694494800114493,
      "grad_norm": 0.2963167726993561,
      "learning_rate": 8.360940727244859e-05,
      "loss": 0.0436,
      "step": 1110
    },
    {
      "epoch": 1.7097605190344431,
      "grad_norm": 0.2605683207511902,
      "learning_rate": 8.321287091572403e-05,
      "loss": 0.0272,
      "step": 1120
    },
    {
      "epoch": 1.7250262379543937,
      "grad_norm": 0.43958792090415955,
      "learning_rate": 8.281256021240566e-05,
      "loss": 0.0317,
      "step": 1130
    },
    {
      "epoch": 1.740291956874344,
      "grad_norm": 0.20269373059272766,
      "learning_rate": 8.240852065423506e-05,
      "loss": 0.0376,
      "step": 1140
    },
    {
      "epoch": 1.7555576757942943,
      "grad_norm": 0.19015586376190186,
      "learning_rate": 8.20007981567048e-05,
      "loss": 0.0318,
      "step": 1150
    },
    {
      "epoch": 1.7555576757942943,
      "eval_loss": 0.04022592678666115,
      "eval_runtime": 156.0262,
      "eval_samples_per_second": 7.467,
      "eval_steps_per_second": 7.467,
      "step": 1150
    },
    {
      "epoch": 1.7708233947142449,
      "grad_norm": 0.2430041879415512,
      "learning_rate": 8.158943905384082e-05,
      "loss": 0.0394,
      "step": 1160
    },
    {
      "epoch": 1.7860891136341952,
      "grad_norm": 0.1872541755437851,
      "learning_rate": 8.117449009293668e-05,
      "loss": 0.031,
      "step": 1170
    },
    {
      "epoch": 1.8013548325541455,
      "grad_norm": 0.24640363454818726,
      "learning_rate": 8.075599842924139e-05,
      "loss": 0.0253,
      "step": 1180
    },
    {
      "epoch": 1.816620551474096,
      "grad_norm": 0.23295819759368896,
      "learning_rate": 8.033401162060049e-05,
      "loss": 0.0362,
      "step": 1190
    },
    {
      "epoch": 1.8318862703940464,
      "grad_norm": 0.5127443671226501,
      "learning_rate": 7.990857762205157e-05,
      "loss": 0.0484,
      "step": 1200
    },
    {
      "epoch": 1.8318862703940464,
      "eval_loss": 0.03849531337618828,
      "eval_runtime": 155.9091,
      "eval_samples_per_second": 7.472,
      "eval_steps_per_second": 7.472,
      "step": 1200
    },
    {
      "epoch": 1.8471519893139967,
      "grad_norm": 0.3558383285999298,
      "learning_rate": 7.947974478037468e-05,
      "loss": 0.0371,
      "step": 1210
    },
    {
      "epoch": 1.8624177082339473,
      "grad_norm": 0.42320486903190613,
      "learning_rate": 7.904756182859797e-05,
      "loss": 0.0297,
      "step": 1220
    },
    {
      "epoch": 1.8776834271538976,
      "grad_norm": 0.16989709436893463,
      "learning_rate": 7.861207788045984e-05,
      "loss": 0.0313,
      "step": 1230
    },
    {
      "epoch": 1.892949146073848,
      "grad_norm": 0.37773561477661133,
      "learning_rate": 7.817334242482738e-05,
      "loss": 0.0328,
      "step": 1240
    },
    {
      "epoch": 1.9082148649937984,
      "grad_norm": 0.1707279086112976,
      "learning_rate": 7.773140532007262e-05,
      "loss": 0.0263,
      "step": 1250
    },
    {
      "epoch": 1.9082148649937984,
      "eval_loss": 0.03550805523991585,
      "eval_runtime": 155.9432,
      "eval_samples_per_second": 7.471,
      "eval_steps_per_second": 7.471,
      "step": 1250
    },
    {
      "epoch": 1.9234805839137485,
      "grad_norm": 0.32426440715789795,
      "learning_rate": 7.728631678840638e-05,
      "loss": 0.0293,
      "step": 1260
    },
    {
      "epoch": 1.938746302833699,
      "grad_norm": 0.23525658249855042,
      "learning_rate": 7.683812741017112e-05,
      "loss": 0.0321,
      "step": 1270
    },
    {
      "epoch": 1.9540120217536494,
      "grad_norm": 0.5105435848236084,
      "learning_rate": 7.638688811809274e-05,
      "loss": 0.0379,
      "step": 1280
    },
    {
      "epoch": 1.9692777406735997,
      "grad_norm": 0.27154791355133057,
      "learning_rate": 7.593265019149275e-05,
      "loss": 0.0279,
      "step": 1290
    },
    {
      "epoch": 1.9845434595935503,
      "grad_norm": 0.41763678193092346,
      "learning_rate": 7.547546525046073e-05,
      "loss": 0.0329,
      "step": 1300
    },
    {
      "epoch": 1.9845434595935503,
      "eval_loss": 0.03486299887299538,
      "eval_runtime": 155.963,
      "eval_samples_per_second": 7.47,
      "eval_steps_per_second": 7.47,
      "step": 1300
    },
    {
      "epoch": 1.9998091785135006,
      "grad_norm": 0.12300687283277512,
      "learning_rate": 7.501538524998812e-05,
      "loss": 0.0365,
      "step": 1310
    },
    {
      "epoch": 2.015074897433451,
      "grad_norm": 0.19598349928855896,
      "learning_rate": 7.455246247406406e-05,
      "loss": 0.0137,
      "step": 1320
    },
    {
      "epoch": 2.0303406163534015,
      "grad_norm": 0.3010258376598358,
      "learning_rate": 7.408674952973382e-05,
      "loss": 0.023,
      "step": 1330
    },
    {
      "epoch": 2.0456063352733516,
      "grad_norm": 0.2672453820705414,
      "learning_rate": 7.361829934112036e-05,
      "loss": 0.0213,
      "step": 1340
    },
    {
      "epoch": 2.060872054193302,
      "grad_norm": 0.21076306700706482,
      "learning_rate": 7.314716514341006e-05,
      "loss": 0.0171,
      "step": 1350
    },
    {
      "epoch": 2.060872054193302,
      "eval_loss": 0.03650743141770363,
      "eval_runtime": 155.864,
      "eval_samples_per_second": 7.474,
      "eval_steps_per_second": 7.474,
      "step": 1350
    },
    {
      "epoch": 2.0761377731132526,
      "grad_norm": 0.2325100302696228,
      "learning_rate": 7.267340047680305e-05,
      "loss": 0.0169,
      "step": 1360
    },
    {
      "epoch": 2.0914034920332027,
      "grad_norm": 0.2855936884880066,
      "learning_rate": 7.21970591804287e-05,
      "loss": 0.0169,
      "step": 1370
    },
    {
      "epoch": 2.1066692109531533,
      "grad_norm": 0.5615317821502686,
      "learning_rate": 7.171819538622747e-05,
      "loss": 0.0227,
      "step": 1380
    },
    {
      "epoch": 2.121934929873104,
      "grad_norm": 0.19267569482326508,
      "learning_rate": 7.123686351279914e-05,
      "loss": 0.0178,
      "step": 1390
    },
    {
      "epoch": 2.137200648793054,
      "grad_norm": 0.2372904121875763,
      "learning_rate": 7.07531182592187e-05,
      "loss": 0.0204,
      "step": 1400
    },
    {
      "epoch": 2.137200648793054,
      "eval_loss": 0.03742413595318794,
      "eval_runtime": 155.8584,
      "eval_samples_per_second": 7.475,
      "eval_steps_per_second": 7.475,
      "step": 1400
    },
    {
      "epoch": 2.1524663677130045,
      "grad_norm": 0.4747006893157959,
      "learning_rate": 7.026701459882026e-05,
      "loss": 0.0294,
      "step": 1410
    },
    {
      "epoch": 2.167732086632955,
      "grad_norm": 0.28667911887168884,
      "learning_rate": 6.977860777294988e-05,
      "loss": 0.0169,
      "step": 1420
    },
    {
      "epoch": 2.182997805552905,
      "grad_norm": 0.28892043232917786,
      "learning_rate": 6.92879532846878e-05,
      "loss": 0.017,
      "step": 1430
    },
    {
      "epoch": 2.1982635244728557,
      "grad_norm": 0.24941062927246094,
      "learning_rate": 6.879510689254104e-05,
      "loss": 0.0182,
      "step": 1440
    },
    {
      "epoch": 2.213529243392806,
      "grad_norm": 0.3995440900325775,
      "learning_rate": 6.830012460410697e-05,
      "loss": 0.0248,
      "step": 1450
    },
    {
      "epoch": 2.213529243392806,
      "eval_loss": 0.04132550209760666,
      "eval_runtime": 155.8248,
      "eval_samples_per_second": 7.476,
      "eval_steps_per_second": 7.476,
      "step": 1450
    },
    {
      "epoch": 2.2287949623127563,
      "grad_norm": 0.2918906509876251,
      "learning_rate": 6.780306266970851e-05,
      "loss": 0.0199,
      "step": 1460
    },
    {
      "epoch": 2.244060681232707,
      "grad_norm": 0.9692716002464294,
      "learning_rate": 6.73039775760018e-05,
      "loss": 0.0166,
      "step": 1470
    },
    {
      "epoch": 2.2593264001526574,
      "grad_norm": 0.2235914021730423,
      "learning_rate": 6.680292603955702e-05,
      "loss": 0.0178,
      "step": 1480
    },
    {
      "epoch": 2.2745921190726075,
      "grad_norm": 0.25725817680358887,
      "learning_rate": 6.629996500041299e-05,
      "loss": 0.0171,
      "step": 1490
    },
    {
      "epoch": 2.289857837992558,
      "grad_norm": 0.1939374804496765,
      "learning_rate": 6.579515161560649e-05,
      "loss": 0.0132,
      "step": 1500
    },
    {
      "epoch": 2.289857837992558,
      "eval_loss": 0.039708200842142105,
      "eval_runtime": 155.7865,
      "eval_samples_per_second": 7.478,
      "eval_steps_per_second": 7.478,
      "step": 1500
    },
    {
      "epoch": 2.305123556912508,
      "grad_norm": 0.5140581727027893,
      "learning_rate": 6.528854325267692e-05,
      "loss": 0.0254,
      "step": 1510
    },
    {
      "epoch": 2.3203892758324587,
      "grad_norm": 0.3529181480407715,
      "learning_rate": 6.478019748314686e-05,
      "loss": 0.0248,
      "step": 1520
    },
    {
      "epoch": 2.3356549947524092,
      "grad_norm": 0.6066356897354126,
      "learning_rate": 6.42701720759797e-05,
      "loss": 0.0229,
      "step": 1530
    },
    {
      "epoch": 2.3509207136723593,
      "grad_norm": 0.13625769317150116,
      "learning_rate": 6.375852499101467e-05,
      "loss": 0.0143,
      "step": 1540
    },
    {
      "epoch": 2.36618643259231,
      "grad_norm": 0.16366352140903473,
      "learning_rate": 6.324531437238019e-05,
      "loss": 0.0152,
      "step": 1550
    },
    {
      "epoch": 2.36618643259231,
      "eval_loss": 0.03828950226306915,
      "eval_runtime": 155.7795,
      "eval_samples_per_second": 7.479,
      "eval_steps_per_second": 7.479,
      "step": 1550
    },
    {
      "epoch": 2.3814521515122604,
      "grad_norm": 0.27981364727020264,
      "learning_rate": 6.273059854188636e-05,
      "loss": 0.0218,
      "step": 1560
    },
    {
      "epoch": 2.3967178704322105,
      "grad_norm": 0.42901352047920227,
      "learning_rate": 6.221443599239721e-05,
      "loss": 0.0179,
      "step": 1570
    },
    {
      "epoch": 2.411983589352161,
      "grad_norm": 0.38229596614837646,
      "learning_rate": 6.169688538118342e-05,
      "loss": 0.0255,
      "step": 1580
    },
    {
      "epoch": 2.4272493082721116,
      "grad_norm": 0.17062246799468994,
      "learning_rate": 6.117800552325655e-05,
      "loss": 0.02,
      "step": 1590
    },
    {
      "epoch": 2.4425150271920617,
      "grad_norm": 0.2703785300254822,
      "learning_rate": 6.06578553846852e-05,
      "loss": 0.0222,
      "step": 1600
    },
    {
      "epoch": 2.4425150271920617,
      "eval_loss": 0.03865242376923561,
      "eval_runtime": 155.7392,
      "eval_samples_per_second": 7.48,
      "eval_steps_per_second": 7.48,
      "step": 1600
    },
    {
      "epoch": 2.4577807461120122,
      "grad_norm": 0.2679692506790161,
      "learning_rate": 6.013649407589401e-05,
      "loss": 0.0223,
      "step": 1610
    },
    {
      "epoch": 2.473046465031963,
      "grad_norm": 0.29233378171920776,
      "learning_rate": 5.961398084494634e-05,
      "loss": 0.0252,
      "step": 1620
    },
    {
      "epoch": 2.488312183951913,
      "grad_norm": 0.46518829464912415,
      "learning_rate": 5.909037507081121e-05,
      "loss": 0.0241,
      "step": 1630
    },
    {
      "epoch": 2.5035779028718634,
      "grad_norm": 0.08025455474853516,
      "learning_rate": 5.8565736256615434e-05,
      "loss": 0.0266,
      "step": 1640
    },
    {
      "epoch": 2.518843621791814,
      "grad_norm": 0.14326392114162445,
      "learning_rate": 5.8040124022881625e-05,
      "loss": 0.0187,
      "step": 1650
    },
    {
      "epoch": 2.518843621791814,
      "eval_loss": 0.03742476552724838,
      "eval_runtime": 155.7547,
      "eval_samples_per_second": 7.48,
      "eval_steps_per_second": 7.48,
      "step": 1650
    },
    {
      "epoch": 2.534109340711764,
      "grad_norm": 0.2520762085914612,
      "learning_rate": 5.751359810075284e-05,
      "loss": 0.0151,
      "step": 1660
    },
    {
      "epoch": 2.5493750596317146,
      "grad_norm": 0.2356204092502594,
      "learning_rate": 5.6986218325204676e-05,
      "loss": 0.0103,
      "step": 1670
    },
    {
      "epoch": 2.5646407785516647,
      "grad_norm": 0.4597303867340088,
      "learning_rate": 5.645804462824556e-05,
      "loss": 0.0151,
      "step": 1680
    },
    {
      "epoch": 2.5799064974716153,
      "grad_norm": 0.40278297662734985,
      "learning_rate": 5.5929137032106005e-05,
      "loss": 0.0171,
      "step": 1690
    },
    {
      "epoch": 2.595172216391566,
      "grad_norm": 0.30628910660743713,
      "learning_rate": 5.53995556424176e-05,
      "loss": 0.0177,
      "step": 1700
    },
    {
      "epoch": 2.595172216391566,
      "eval_loss": 0.041493531316518784,
      "eval_runtime": 155.6991,
      "eval_samples_per_second": 7.482,
      "eval_steps_per_second": 7.482,
      "step": 1700
    },
    {
      "epoch": 2.610437935311516,
      "grad_norm": 0.2515423595905304,
      "learning_rate": 5.4869360641382615e-05,
      "loss": 0.0153,
      "step": 1710
    },
    {
      "epoch": 2.6257036542314665,
      "grad_norm": 0.36887672543525696,
      "learning_rate": 5.433861228093471e-05,
      "loss": 0.023,
      "step": 1720
    },
    {
      "epoch": 2.640969373151417,
      "grad_norm": 0.19568361341953278,
      "learning_rate": 5.380737087589197e-05,
      "loss": 0.0144,
      "step": 1730
    },
    {
      "epoch": 2.656235092071367,
      "grad_norm": 0.476130872964859,
      "learning_rate": 5.327569679710256e-05,
      "loss": 0.0286,
      "step": 1740
    },
    {
      "epoch": 2.6715008109913176,
      "grad_norm": 0.26536592841148376,
      "learning_rate": 5.274365046458416e-05,
      "loss": 0.0164,
      "step": 1750
    },
    {
      "epoch": 2.6715008109913176,
      "eval_loss": 0.03795096278190613,
      "eval_runtime": 155.706,
      "eval_samples_per_second": 7.482,
      "eval_steps_per_second": 7.482,
      "step": 1750
    },
    {
      "epoch": 2.686766529911268,
      "grad_norm": 0.5614975690841675,
      "learning_rate": 5.2211292340657804e-05,
      "loss": 0.0187,
      "step": 1760
    },
    {
      "epoch": 2.7020322488312183,
      "grad_norm": 0.1835988163948059,
      "learning_rate": 5.167868292307678e-05,
      "loss": 0.0242,
      "step": 1770
    },
    {
      "epoch": 2.717297967751169,
      "grad_norm": 0.2698806822299957,
      "learning_rate": 5.114588273815173e-05,
      "loss": 0.016,
      "step": 1780
    },
    {
      "epoch": 2.732563686671119,
      "grad_norm": 0.5950194597244263,
      "learning_rate": 5.061295233387223e-05,
      "loss": 0.0275,
      "step": 1790
    },
    {
      "epoch": 2.7478294055910695,
      "grad_norm": 0.2528298795223236,
      "learning_rate": 5.007995227302617e-05,
      "loss": 0.0212,
      "step": 1800
    },
    {
      "epoch": 2.7478294055910695,
      "eval_loss": 0.03947535157203674,
      "eval_runtime": 155.725,
      "eval_samples_per_second": 7.481,
      "eval_steps_per_second": 7.481,
      "step": 1800
    },
    {
      "epoch": 2.76309512451102,
      "grad_norm": 0.09799537807703018,
      "learning_rate": 4.954694312631729e-05,
      "loss": 0.024,
      "step": 1810
    },
    {
      "epoch": 2.7783608434309706,
      "grad_norm": 0.6346628665924072,
      "learning_rate": 4.901398546548181e-05,
      "loss": 0.0186,
      "step": 1820
    },
    {
      "epoch": 2.7936265623509207,
      "grad_norm": 0.4291705787181854,
      "learning_rate": 4.848113985640513e-05,
      "loss": 0.0178,
      "step": 1830
    },
    {
      "epoch": 2.808892281270871,
      "grad_norm": 0.26121002435684204,
      "learning_rate": 4.794846685223884e-05,
      "loss": 0.0229,
      "step": 1840
    },
    {
      "epoch": 2.8241580001908213,
      "grad_norm": 0.40268027782440186,
      "learning_rate": 4.741602698651966e-05,
      "loss": 0.0248,
      "step": 1850
    },
    {
      "epoch": 2.8241580001908213,
      "eval_loss": 0.03567200526595116,
      "eval_runtime": 155.7614,
      "eval_samples_per_second": 7.479,
      "eval_steps_per_second": 7.479,
      "step": 1850
    },
    {
      "epoch": 2.839423719110772,
      "grad_norm": 0.28412625193595886,
      "learning_rate": 4.6883880766290086e-05,
      "loss": 0.0244,
      "step": 1860
    },
    {
      "epoch": 2.8546894380307224,
      "grad_norm": 0.2994270324707031,
      "learning_rate": 4.635208866522251e-05,
      "loss": 0.0176,
      "step": 1870
    },
    {
      "epoch": 2.8699551569506725,
      "grad_norm": 0.3112628757953644,
      "learning_rate": 4.5820711116746785e-05,
      "loss": 0.0136,
      "step": 1880
    },
    {
      "epoch": 2.885220875870623,
      "grad_norm": 0.7493220567703247,
      "learning_rate": 4.528980850718255e-05,
      "loss": 0.0249,
      "step": 1890
    },
    {
      "epoch": 2.9004865947905736,
      "grad_norm": 0.3043513894081116,
      "learning_rate": 4.475944116887695e-05,
      "loss": 0.0187,
      "step": 1900
    },
    {
      "epoch": 2.9004865947905736,
      "eval_loss": 0.03836146742105484,
      "eval_runtime": 155.7298,
      "eval_samples_per_second": 7.481,
      "eval_steps_per_second": 7.481,
      "step": 1900
    },
    {
      "epoch": 2.9157523137105237,
      "grad_norm": 0.28810712695121765,
      "learning_rate": 4.4229669373348226e-05,
      "loss": 0.0252,
      "step": 1910
    },
    {
      "epoch": 2.9310180326304742,
      "grad_norm": 0.26556193828582764,
      "learning_rate": 4.3700553324436575e-05,
      "loss": 0.0177,
      "step": 1920
    },
    {
      "epoch": 2.9462837515504248,
      "grad_norm": 0.4432234764099121,
      "learning_rate": 4.317215315146238e-05,
      "loss": 0.0247,
      "step": 1930
    },
    {
      "epoch": 2.961549470470375,
      "grad_norm": 0.35339197516441345,
      "learning_rate": 4.264452890239315e-05,
      "loss": 0.0217,
      "step": 1940
    },
    {
      "epoch": 2.9768151893903254,
      "grad_norm": 0.48235374689102173,
      "learning_rate": 4.211774053701952e-05,
      "loss": 0.0315,
      "step": 1950
    },
    {
      "epoch": 2.9768151893903254,
      "eval_loss": 0.037155311554670334,
      "eval_runtime": 155.6439,
      "eval_samples_per_second": 7.485,
      "eval_steps_per_second": 7.485,
      "step": 1950
    },
    {
      "epoch": 2.9920809083102755,
      "grad_norm": 0.1263391226530075,
      "learning_rate": 4.159184792014145e-05,
      "loss": 0.0153,
      "step": 1960
    },
    {
      "epoch": 3.007346627230226,
      "grad_norm": 0.1030694767832756,
      "learning_rate": 4.1066910814765016e-05,
      "loss": 0.0102,
      "step": 1970
    },
    {
      "epoch": 3.0226123461501766,
      "grad_norm": 0.19300442934036255,
      "learning_rate": 4.0542988875310995e-05,
      "loss": 0.0146,
      "step": 1980
    },
    {
      "epoch": 3.0378780650701267,
      "grad_norm": 0.14339104294776917,
      "learning_rate": 4.002014164083552e-05,
      "loss": 0.0098,
      "step": 1990
    },
    {
      "epoch": 3.0531437839900772,
      "grad_norm": 0.06706851720809937,
      "learning_rate": 3.9498428528264204e-05,
      "loss": 0.006,
      "step": 2000
    },
    {
      "epoch": 3.0531437839900772,
      "eval_loss": 0.04230096563696861,
      "eval_runtime": 155.6887,
      "eval_samples_per_second": 7.483,
      "eval_steps_per_second": 7.483,
      "step": 2000
    },
    {
      "epoch": 3.068409502910028,
      "grad_norm": 0.12648242712020874,
      "learning_rate": 3.89779088256397e-05,
      "loss": 0.0063,
      "step": 2010
    },
    {
      "epoch": 3.083675221829978,
      "grad_norm": 0.24645593762397766,
      "learning_rate": 3.845864168538437e-05,
      "loss": 0.0067,
      "step": 2020
    },
    {
      "epoch": 3.0989409407499284,
      "grad_norm": 0.23762091994285583,
      "learning_rate": 3.794068611757794e-05,
      "loss": 0.0094,
      "step": 2030
    },
    {
      "epoch": 3.114206659669879,
      "grad_norm": 0.22739964723587036,
      "learning_rate": 3.7424100983251695e-05,
      "loss": 0.0094,
      "step": 2040
    },
    {
      "epoch": 3.129472378589829,
      "grad_norm": 0.7959986925125122,
      "learning_rate": 3.6908944987699345e-05,
      "loss": 0.0077,
      "step": 2050
    },
    {
      "epoch": 3.129472378589829,
      "eval_loss": 0.04588808864355087,
      "eval_runtime": 155.6391,
      "eval_samples_per_second": 7.485,
      "eval_steps_per_second": 7.485,
      "step": 2050
    },
    {
      "epoch": 3.1447380975097796,
      "grad_norm": 0.09930922836065292,
      "learning_rate": 3.639527667380571e-05,
      "loss": 0.0035,
      "step": 2060
    },
    {
      "epoch": 3.16000381642973,
      "grad_norm": 0.3177180290222168,
      "learning_rate": 3.5883154415393885e-05,
      "loss": 0.0079,
      "step": 2070
    },
    {
      "epoch": 3.1752695353496803,
      "grad_norm": 0.402797132730484,
      "learning_rate": 3.537263641059152e-05,
      "loss": 0.0098,
      "step": 2080
    },
    {
      "epoch": 3.190535254269631,
      "grad_norm": 0.27361375093460083,
      "learning_rate": 3.486378067521718e-05,
      "loss": 0.0087,
      "step": 2090
    },
    {
      "epoch": 3.2058009731895813,
      "grad_norm": 0.41781383752822876,
      "learning_rate": 3.435664503618732e-05,
      "loss": 0.0073,
      "step": 2100
    },
    {
      "epoch": 3.2058009731895813,
      "eval_loss": 0.04926072433590889,
      "eval_runtime": 155.69,
      "eval_samples_per_second": 7.483,
      "eval_steps_per_second": 7.483,
      "step": 2100
    },
    {
      "epoch": 3.2210666921095314,
      "grad_norm": 0.04020585119724274,
      "learning_rate": 3.3851287124944756e-05,
      "loss": 0.0059,
      "step": 2110
    },
    {
      "epoch": 3.236332411029482,
      "grad_norm": 0.11026966571807861,
      "learning_rate": 3.334776437090944e-05,
      "loss": 0.0052,
      "step": 2120
    },
    {
      "epoch": 3.251598129949432,
      "grad_norm": 0.029121991246938705,
      "learning_rate": 3.2846133994952046e-05,
      "loss": 0.0044,
      "step": 2130
    },
    {
      "epoch": 3.2668638488693826,
      "grad_norm": 0.22142323851585388,
      "learning_rate": 3.234645300289136e-05,
      "loss": 0.0039,
      "step": 2140
    },
    {
      "epoch": 3.282129567789333,
      "grad_norm": 0.9102628827095032,
      "learning_rate": 3.1848778179016075e-05,
      "loss": 0.0096,
      "step": 2150
    },
    {
      "epoch": 3.282129567789333,
      "eval_loss": 0.05229068547487259,
      "eval_runtime": 155.7655,
      "eval_samples_per_second": 7.479,
      "eval_steps_per_second": 7.479,
      "step": 2150
    },
    {
      "epoch": 3.2973952867092833,
      "grad_norm": 0.6087216138839722,
      "learning_rate": 3.135316607963176e-05,
      "loss": 0.0104,
      "step": 2160
    },
    {
      "epoch": 3.312661005629234,
      "grad_norm": 0.49881669878959656,
      "learning_rate": 3.085967302663375e-05,
      "loss": 0.0076,
      "step": 2170
    },
    {
      "epoch": 3.3279267245491844,
      "grad_norm": 0.3587443232536316,
      "learning_rate": 3.0368355101106615e-05,
      "loss": 0.0142,
      "step": 2180
    },
    {
      "epoch": 3.3431924434691345,
      "grad_norm": 0.12659725546836853,
      "learning_rate": 2.987926813695116e-05,
      "loss": 0.008,
      "step": 2190
    },
    {
      "epoch": 3.358458162389085,
      "grad_norm": 0.07203304767608643,
      "learning_rate": 2.939246771453924e-05,
      "loss": 0.0086,
      "step": 2200
    },
    {
      "epoch": 3.358458162389085,
      "eval_loss": 0.04494870454072952,
      "eval_runtime": 155.7823,
      "eval_samples_per_second": 7.478,
      "eval_steps_per_second": 7.478,
      "step": 2200
    },
    {
      "epoch": 3.3737238813090356,
      "grad_norm": 0.7683483958244324,
      "learning_rate": 2.890800915439772e-05,
      "loss": 0.0112,
      "step": 2210
    },
    {
      "epoch": 3.3889896002289857,
      "grad_norm": 0.1044091284275055,
      "learning_rate": 2.842594751092159e-05,
      "loss": 0.01,
      "step": 2220
    },
    {
      "epoch": 3.404255319148936,
      "grad_norm": 0.2757960557937622,
      "learning_rate": 2.794633756611776e-05,
      "loss": 0.0082,
      "step": 2230
    },
    {
      "epoch": 3.4195210380688867,
      "grad_norm": 0.31744205951690674,
      "learning_rate": 2.7469233823379347e-05,
      "loss": 0.0056,
      "step": 2240
    },
    {
      "epoch": 3.434786756988837,
      "grad_norm": 1.4233782291412354,
      "learning_rate": 2.6994690501292032e-05,
      "loss": 0.0057,
      "step": 2250
    },
    {
      "epoch": 3.434786756988837,
      "eval_loss": 0.04692812263965607,
      "eval_runtime": 155.7735,
      "eval_samples_per_second": 7.479,
      "eval_steps_per_second": 7.479,
      "step": 2250
    },
    {
      "epoch": 3.4500524759087874,
      "grad_norm": 0.5911946892738342,
      "learning_rate": 2.652276152747246e-05,
      "loss": 0.0073,
      "step": 2260
    },
    {
      "epoch": 3.465318194828738,
      "grad_norm": 0.26131606101989746,
      "learning_rate": 2.6053500532439968e-05,
      "loss": 0.0056,
      "step": 2270
    },
    {
      "epoch": 3.480583913748688,
      "grad_norm": 0.03926728665828705,
      "learning_rate": 2.5586960843521824e-05,
      "loss": 0.0085,
      "step": 2280
    },
    {
      "epoch": 3.4958496326686386,
      "grad_norm": 0.3418906629085541,
      "learning_rate": 2.5123195478793217e-05,
      "loss": 0.0096,
      "step": 2290
    },
    {
      "epoch": 3.5111153515885887,
      "grad_norm": 0.1729433536529541,
      "learning_rate": 2.466225714105202e-05,
      "loss": 0.0098,
      "step": 2300
    },
    {
      "epoch": 3.5111153515885887,
      "eval_loss": 0.04600750282406807,
      "eval_runtime": 155.6908,
      "eval_samples_per_second": 7.483,
      "eval_steps_per_second": 7.483,
      "step": 2300
    },
    {
      "epoch": 3.526381070508539,
      "grad_norm": 0.1316080242395401,
      "learning_rate": 2.420419821182982e-05,
      "loss": 0.0072,
      "step": 2310
    },
    {
      "epoch": 3.5416467894284898,
      "grad_norm": 0.6513664722442627,
      "learning_rate": 2.3749070745438996e-05,
      "loss": 0.0056,
      "step": 2320
    },
    {
      "epoch": 3.5569125083484403,
      "grad_norm": 0.20830810070037842,
      "learning_rate": 2.3296926463057396e-05,
      "loss": 0.0071,
      "step": 2330
    },
    {
      "epoch": 3.5721782272683904,
      "grad_norm": 0.14865128695964813,
      "learning_rate": 2.284781674685058e-05,
      "loss": 0.0041,
      "step": 2340
    },
    {
      "epoch": 3.587443946188341,
      "grad_norm": 0.1634235978126526,
      "learning_rate": 2.2401792634132708e-05,
      "loss": 0.0086,
      "step": 2350
    },
    {
      "epoch": 3.587443946188341,
      "eval_loss": 0.04932467266917229,
      "eval_runtime": 155.7011,
      "eval_samples_per_second": 7.482,
      "eval_steps_per_second": 7.482,
      "step": 2350
    },
    {
      "epoch": 3.602709665108291,
      "grad_norm": 0.6523926854133606,
      "learning_rate": 2.195890481156666e-05,
      "loss": 0.0083,
      "step": 2360
    },
    {
      "epoch": 3.6179753840282416,
      "grad_norm": 0.15616604685783386,
      "learning_rate": 2.151920360940387e-05,
      "loss": 0.0106,
      "step": 2370
    },
    {
      "epoch": 3.633241102948192,
      "grad_norm": 0.43568360805511475,
      "learning_rate": 2.1082738995764785e-05,
      "loss": 0.0052,
      "step": 2380
    },
    {
      "epoch": 3.6485068218681422,
      "grad_norm": 0.251506507396698,
      "learning_rate": 2.0649560570960465e-05,
      "loss": 0.0056,
      "step": 2390
    },
    {
      "epoch": 3.663772540788093,
      "grad_norm": 0.3159063458442688,
      "learning_rate": 2.0219717561855855e-05,
      "loss": 0.0073,
      "step": 2400
    },
    {
      "epoch": 3.663772540788093,
      "eval_loss": 0.047119684517383575,
      "eval_runtime": 155.735,
      "eval_samples_per_second": 7.481,
      "eval_steps_per_second": 7.481,
      "step": 2400
    },
    {
      "epoch": 3.679038259708043,
      "grad_norm": 0.048487480729818344,
      "learning_rate": 1.9793258816275728e-05,
      "loss": 0.0091,
      "step": 2410
    },
    {
      "epoch": 3.6943039786279934,
      "grad_norm": 0.4514082968235016,
      "learning_rate": 1.9370232797453402e-05,
      "loss": 0.009,
      "step": 2420
    },
    {
      "epoch": 3.709569697547944,
      "grad_norm": 0.271023154258728,
      "learning_rate": 1.8950687578523502e-05,
      "loss": 0.0062,
      "step": 2430
    },
    {
      "epoch": 3.7248354164678945,
      "grad_norm": 0.155420184135437,
      "learning_rate": 1.853467083705869e-05,
      "loss": 0.0031,
      "step": 2440
    },
    {
      "epoch": 3.7401011353878446,
      "grad_norm": 0.33148664236068726,
      "learning_rate": 1.8122229849651716e-05,
      "loss": 0.0086,
      "step": 2450
    },
    {
      "epoch": 3.7401011353878446,
      "eval_loss": 0.04678386077284813,
      "eval_runtime": 155.695,
      "eval_samples_per_second": 7.483,
      "eval_steps_per_second": 7.483,
      "step": 2450
    },
    {
      "epoch": 3.755366854307795,
      "grad_norm": 0.3089897632598877,
      "learning_rate": 1.7713411486542707e-05,
      "loss": 0.008,
      "step": 2460
    },
    {
      "epoch": 3.7706325732277453,
      "grad_norm": 0.4456401467323303,
      "learning_rate": 1.7308262206292897e-05,
      "loss": 0.0083,
      "step": 2470
    },
    {
      "epoch": 3.785898292147696,
      "grad_norm": 0.46226081252098083,
      "learning_rate": 1.6906828050504907e-05,
      "loss": 0.0149,
      "step": 2480
    },
    {
      "epoch": 3.8011640110676463,
      "grad_norm": 0.018434859812259674,
      "learning_rate": 1.650915463859068e-05,
      "loss": 0.0077,
      "step": 2490
    },
    {
      "epoch": 3.8164297299875964,
      "grad_norm": 0.28657904267311096,
      "learning_rate": 1.6115287162587055e-05,
      "loss": 0.0079,
      "step": 2500
    },
    {
      "epoch": 3.8164297299875964,
      "eval_loss": 0.045529987663030624,
      "eval_runtime": 155.7642,
      "eval_samples_per_second": 7.479,
      "eval_steps_per_second": 7.479,
      "step": 2500
    },
    {
      "epoch": 3.831695448907547,
      "grad_norm": 0.09848945587873459,
      "learning_rate": 1.57252703820203e-05,
      "loss": 0.0052,
      "step": 2510
    },
    {
      "epoch": 3.8469611678274975,
      "grad_norm": 0.4062275290489197,
      "learning_rate": 1.5339148618819393e-05,
      "loss": 0.009,
      "step": 2520
    },
    {
      "epoch": 3.8622268867474476,
      "grad_norm": 0.3156788945198059,
      "learning_rate": 1.4956965752279395e-05,
      "loss": 0.0082,
      "step": 2530
    },
    {
      "epoch": 3.877492605667398,
      "grad_norm": 0.35523831844329834,
      "learning_rate": 1.457876521407484e-05,
      "loss": 0.0072,
      "step": 2540
    },
    {
      "epoch": 3.8927583245873487,
      "grad_norm": 0.243367001414299,
      "learning_rate": 1.4204589983324173e-05,
      "loss": 0.0042,
      "step": 2550
    },
    {
      "epoch": 3.8927583245873487,
      "eval_loss": 0.04733826965093613,
      "eval_runtime": 155.9113,
      "eval_samples_per_second": 7.472,
      "eval_steps_per_second": 7.472,
      "step": 2550
    },
    {
      "epoch": 3.908024043507299,
      "grad_norm": 0.5131608843803406,
      "learning_rate": 1.383448258170557e-05,
      "loss": 0.0091,
      "step": 2560
    },
    {
      "epoch": 3.9232897624272494,
      "grad_norm": 0.4350660443305969,
      "learning_rate": 1.3468485068624653e-05,
      "loss": 0.005,
      "step": 2570
    },
    {
      "epoch": 3.9385554813471995,
      "grad_norm": 0.10545644909143448,
      "learning_rate": 1.310663903643492e-05,
      "loss": 0.009,
      "step": 2580
    },
    {
      "epoch": 3.95382120026715,
      "grad_norm": 0.19669237732887268,
      "learning_rate": 1.2748985605711028e-05,
      "loss": 0.0072,
      "step": 2590
    },
    {
      "epoch": 3.9690869191871005,
      "grad_norm": 0.3060455024242401,
      "learning_rate": 1.2395565420575932e-05,
      "loss": 0.0088,
      "step": 2600
    },
    {
      "epoch": 3.9690869191871005,
      "eval_loss": 0.047428667545318604,
      "eval_runtime": 156.1393,
      "eval_samples_per_second": 7.461,
      "eval_steps_per_second": 7.461,
      "step": 2600
    },
    {
      "epoch": 3.984352638107051,
      "grad_norm": 0.1554788053035736,
      "learning_rate": 1.2046418644081903e-05,
      "loss": 0.014,
      "step": 2610
    },
    {
      "epoch": 3.999618357027001,
      "grad_norm": 1.1880522966384888,
      "learning_rate": 1.1701584953646505e-05,
      "loss": 0.0068,
      "step": 2620
    },
    {
      "epoch": 4.014884075946951,
      "grad_norm": 0.056341659277677536,
      "learning_rate": 1.1361103536543466e-05,
      "loss": 0.0044,
      "step": 2630
    },
    {
      "epoch": 4.030149794866902,
      "grad_norm": 0.012919405475258827,
      "learning_rate": 1.1025013085449527e-05,
      "loss": 0.0022,
      "step": 2640
    },
    {
      "epoch": 4.045415513786852,
      "grad_norm": 0.060808297246694565,
      "learning_rate": 1.0693351794047224e-05,
      "loss": 0.004,
      "step": 2650
    },
    {
      "epoch": 4.045415513786852,
      "eval_loss": 0.0473913699388504,
      "eval_runtime": 156.2274,
      "eval_samples_per_second": 7.457,
      "eval_steps_per_second": 7.457,
      "step": 2650
    },
    {
      "epoch": 4.060681232706803,
      "grad_norm": 0.15758514404296875,
      "learning_rate": 1.036615735268468e-05,
      "loss": 0.0015,
      "step": 2660
    },
    {
      "epoch": 4.0759469516267535,
      "grad_norm": 0.04278062283992767,
      "learning_rate": 1.0043466944092272e-05,
      "loss": 0.0023,
      "step": 2670
    },
    {
      "epoch": 4.091212670546703,
      "grad_norm": 0.08595144748687744,
      "learning_rate": 9.72531723915726e-06,
      "loss": 0.0048,
      "step": 2680
    },
    {
      "epoch": 4.106478389466654,
      "grad_norm": 0.03355566784739494,
      "learning_rate": 9.411744392756405e-06,
      "loss": 0.0033,
      "step": 2690
    },
    {
      "epoch": 4.121744108386604,
      "grad_norm": 0.07282890379428864,
      "learning_rate": 9.102784039647339e-06,
      "loss": 0.0033,
      "step": 2700
    },
    {
      "epoch": 4.121744108386604,
      "eval_loss": 0.049959179013967514,
      "eval_runtime": 156.1823,
      "eval_samples_per_second": 7.459,
      "eval_steps_per_second": 7.459,
      "step": 2700
    },
    {
      "epoch": 4.137009827306555,
      "grad_norm": 0.01570916548371315,
      "learning_rate": 8.79847129041893e-06,
      "loss": 0.003,
      "step": 2710
    },
    {
      "epoch": 4.152275546226505,
      "grad_norm": 0.06394907087087631,
      "learning_rate": 8.498840727501316e-06,
      "loss": 0.0019,
      "step": 2720
    },
    {
      "epoch": 4.167541265146456,
      "grad_norm": 0.10380612313747406,
      "learning_rate": 8.203926401235957e-06,
      "loss": 0.0034,
      "step": 2730
    },
    {
      "epoch": 4.1828069840664055,
      "grad_norm": 0.1005665585398674,
      "learning_rate": 7.913761826006017e-06,
      "loss": 0.0023,
      "step": 2740
    },
    {
      "epoch": 4.198072702986356,
      "grad_norm": 0.058339644223451614,
      "learning_rate": 7.628379976427868e-06,
      "loss": 0.0009,
      "step": 2750
    },
    {
      "epoch": 4.198072702986356,
      "eval_loss": 0.05198967829346657,
      "eval_runtime": 156.1523,
      "eval_samples_per_second": 7.461,
      "eval_steps_per_second": 7.461,
      "step": 2750
    },
    {
      "epoch": 4.213338421906307,
      "grad_norm": 0.1591460257768631,
      "learning_rate": 7.347813283603705e-06,
      "loss": 0.0058,
      "step": 2760
    },
    {
      "epoch": 4.228604140826257,
      "grad_norm": 0.017265984788537025,
      "learning_rate": 7.072093631436161e-06,
      "loss": 0.0014,
      "step": 2770
    },
    {
      "epoch": 4.243869859746208,
      "grad_norm": 0.05040580779314041,
      "learning_rate": 6.801252353004867e-06,
      "loss": 0.0033,
      "step": 2780
    },
    {
      "epoch": 4.259135578666158,
      "grad_norm": 0.04195038229227066,
      "learning_rate": 6.535320227005826e-06,
      "loss": 0.0034,
      "step": 2790
    },
    {
      "epoch": 4.274401297586108,
      "grad_norm": 0.024743085727095604,
      "learning_rate": 6.274327474253611e-06,
      "loss": 0.0021,
      "step": 2800
    },
    {
      "epoch": 4.274401297586108,
      "eval_loss": 0.053240980952978134,
      "eval_runtime": 156.2448,
      "eval_samples_per_second": 7.456,
      "eval_steps_per_second": 7.456,
      "step": 2800
    },
    {
      "epoch": 4.289667016506058,
      "grad_norm": 0.03961138427257538,
      "learning_rate": 6.018303754247112e-06,
      "loss": 0.0027,
      "step": 2810
    },
    {
      "epoch": 4.304932735426009,
      "grad_norm": 0.036531973630189896,
      "learning_rate": 5.767278161798911e-06,
      "loss": 0.0034,
      "step": 2820
    },
    {
      "epoch": 4.3201984543459595,
      "grad_norm": 0.16394208371639252,
      "learning_rate": 5.521279223729026e-06,
      "loss": 0.0012,
      "step": 2830
    },
    {
      "epoch": 4.33546417326591,
      "grad_norm": 0.032556578516960144,
      "learning_rate": 5.280334895622968e-06,
      "loss": 0.0014,
      "step": 2840
    },
    {
      "epoch": 4.350729892185861,
      "grad_norm": 0.21047158539295197,
      "learning_rate": 5.044472558654961e-06,
      "loss": 0.0042,
      "step": 2850
    },
    {
      "epoch": 4.350729892185861,
      "eval_loss": 0.05455264449119568,
      "eval_runtime": 156.3026,
      "eval_samples_per_second": 7.453,
      "eval_steps_per_second": 7.453,
      "step": 2850
    },
    {
      "epoch": 4.36599561110581,
      "grad_norm": 0.02670607902109623,
      "learning_rate": 4.813719016476203e-06,
      "loss": 0.0021,
      "step": 2860
    },
    {
      "epoch": 4.381261330025761,
      "grad_norm": 0.023184500634670258,
      "learning_rate": 4.588100492168973e-06,
      "loss": 0.0012,
      "step": 2870
    },
    {
      "epoch": 4.396527048945711,
      "grad_norm": 0.14299553632736206,
      "learning_rate": 4.367642625266511e-06,
      "loss": 0.0015,
      "step": 2880
    },
    {
      "epoch": 4.411792767865662,
      "grad_norm": 0.020235197618603706,
      "learning_rate": 4.1523704688394176e-06,
      "loss": 0.0014,
      "step": 2890
    },
    {
      "epoch": 4.427058486785612,
      "grad_norm": 0.16960173845291138,
      "learning_rate": 3.9423084866484884e-06,
      "loss": 0.0017,
      "step": 2900
    },
    {
      "epoch": 4.427058486785612,
      "eval_loss": 0.05663292482495308,
      "eval_runtime": 156.2522,
      "eval_samples_per_second": 7.456,
      "eval_steps_per_second": 7.456,
      "step": 2900
    },
    {
      "epoch": 4.442324205705562,
      "grad_norm": 0.030367441475391388,
      "learning_rate": 3.737480550364736e-06,
      "loss": 0.0023,
      "step": 2910
    },
    {
      "epoch": 4.457589924625513,
      "grad_norm": 0.05753505975008011,
      "learning_rate": 3.5379099368564817e-06,
      "loss": 0.0016,
      "step": 2920
    },
    {
      "epoch": 4.472855643545463,
      "grad_norm": 0.1254509538412094,
      "learning_rate": 3.3436193255442396e-06,
      "loss": 0.0013,
      "step": 2930
    },
    {
      "epoch": 4.488121362465414,
      "grad_norm": 0.09961813688278198,
      "learning_rate": 3.1546307958233214e-06,
      "loss": 0.0043,
      "step": 2940
    },
    {
      "epoch": 4.503387081385364,
      "grad_norm": 0.1321631669998169,
      "learning_rate": 2.9709658245547834e-06,
      "loss": 0.002,
      "step": 2950
    },
    {
      "epoch": 4.503387081385364,
      "eval_loss": 0.057674556970596313,
      "eval_runtime": 156.1774,
      "eval_samples_per_second": 7.459,
      "eval_steps_per_second": 7.459,
      "step": 2950
    },
    {
      "epoch": 4.518652800305315,
      "grad_norm": 0.04699141904711723,
      "learning_rate": 2.792645283624712e-06,
      "loss": 0.0022,
      "step": 2960
    },
    {
      "epoch": 4.5339185192252645,
      "grad_norm": 0.13197575509548187,
      "learning_rate": 2.6196894375723645e-06,
      "loss": 0.0019,
      "step": 2970
    },
    {
      "epoch": 4.549184238145215,
      "grad_norm": 0.18447992205619812,
      "learning_rate": 2.452117941287246e-06,
      "loss": 0.0036,
      "step": 2980
    },
    {
      "epoch": 4.5644499570651655,
      "grad_norm": 0.26583021879196167,
      "learning_rate": 2.2899498377755566e-06,
      "loss": 0.0011,
      "step": 2990
    },
    {
      "epoch": 4.579715675985116,
      "grad_norm": 0.02824637107551098,
      "learning_rate": 2.1332035559960663e-06,
      "loss": 0.0021,
      "step": 3000
    },
    {
      "epoch": 4.579715675985116,
      "eval_loss": 0.05825762450695038,
      "eval_runtime": 156.2544,
      "eval_samples_per_second": 7.456,
      "eval_steps_per_second": 7.456,
      "step": 3000
    },
    {
      "epoch": 4.594981394905067,
      "grad_norm": 0.050509508699178696,
      "learning_rate": 1.9818969087658735e-06,
      "loss": 0.0014,
      "step": 3010
    },
    {
      "epoch": 4.610247113825016,
      "grad_norm": 0.05490247160196304,
      "learning_rate": 1.8360470907361093e-06,
      "loss": 0.0011,
      "step": 3020
    },
    {
      "epoch": 4.625512832744967,
      "grad_norm": 0.03532513976097107,
      "learning_rate": 1.6956706764379438e-06,
      "loss": 0.0022,
      "step": 3030
    },
    {
      "epoch": 4.640778551664917,
      "grad_norm": 0.19588930904865265,
      "learning_rate": 1.5607836183989921e-06,
      "loss": 0.0014,
      "step": 3040
    },
    {
      "epoch": 4.656044270584868,
      "grad_norm": 0.04761837422847748,
      "learning_rate": 1.4314012453305215e-06,
      "loss": 0.0017,
      "step": 3050
    },
    {
      "epoch": 4.656044270584868,
      "eval_loss": 0.05846200883388519,
      "eval_runtime": 156.1503,
      "eval_samples_per_second": 7.461,
      "eval_steps_per_second": 7.461,
      "step": 3050
    },
    {
      "epoch": 4.6713099895048185,
      "grad_norm": 0.4008547365665436,
      "learning_rate": 1.3075382603854157e-06,
      "loss": 0.0014,
      "step": 3060
    },
    {
      "epoch": 4.686575708424769,
      "grad_norm": 0.026124007999897003,
      "learning_rate": 1.1892087394873353e-06,
      "loss": 0.0015,
      "step": 3070
    },
    {
      "epoch": 4.701841427344719,
      "grad_norm": 0.1011100560426712,
      "learning_rate": 1.076426129731084e-06,
      "loss": 0.0022,
      "step": 3080
    },
    {
      "epoch": 4.717107146264669,
      "grad_norm": 0.09242038428783417,
      "learning_rate": 9.692032478545e-07,
      "loss": 0.002,
      "step": 3090
    },
    {
      "epoch": 4.73237286518462,
      "grad_norm": 0.017347002401947975,
      "learning_rate": 8.675522787819023e-07,
      "loss": 0.0036,
      "step": 3100
    },
    {
      "epoch": 4.73237286518462,
      "eval_loss": 0.058684661984443665,
      "eval_runtime": 156.2813,
      "eval_samples_per_second": 7.455,
      "eval_steps_per_second": 7.455,
      "step": 3100
    },
    {
      "epoch": 4.74763858410457,
      "grad_norm": 0.20832569897174835,
      "learning_rate": 7.714847742394337e-07,
      "loss": 0.0035,
      "step": 3110
    },
    {
      "epoch": 4.762904303024521,
      "grad_norm": 0.20151978731155396,
      "learning_rate": 6.810116514422593e-07,
      "loss": 0.0046,
      "step": 3120
    },
    {
      "epoch": 4.7781700219444705,
      "grad_norm": 0.048032715916633606,
      "learning_rate": 5.961431918539817e-07,
      "loss": 0.0025,
      "step": 3130
    },
    {
      "epoch": 4.793435740864421,
      "grad_norm": 0.013832991011440754,
      "learning_rate": 5.16889040018187e-07,
      "loss": 0.002,
      "step": 3140
    },
    {
      "epoch": 4.808701459784372,
      "grad_norm": 0.12184188514947891,
      "learning_rate": 4.432582024624543e-07,
      "loss": 0.0021,
      "step": 3150
    },
    {
      "epoch": 4.808701459784372,
      "eval_loss": 0.05884731933474541,
      "eval_runtime": 156.187,
      "eval_samples_per_second": 7.459,
      "eval_steps_per_second": 7.459,
      "step": 3150
    },
    {
      "epoch": 4.823967178704322,
      "grad_norm": 0.14583688974380493,
      "learning_rate": 3.7525904667484737e-07,
      "loss": 0.0015,
      "step": 3160
    },
    {
      "epoch": 4.839232897624273,
      "grad_norm": 1.1857644319534302,
      "learning_rate": 3.128993001530245e-07,
      "loss": 0.0025,
      "step": 3170
    },
    {
      "epoch": 4.854498616544223,
      "grad_norm": 0.10287916660308838,
      "learning_rate": 2.5618604952605816e-07,
      "loss": 0.003,
      "step": 3180
    },
    {
      "epoch": 4.869764335464174,
      "grad_norm": 0.02538507618010044,
      "learning_rate": 2.0512573974912908e-07,
      "loss": 0.0024,
      "step": 3190
    },
    {
      "epoch": 4.885030054384123,
      "grad_norm": 0.5639934539794922,
      "learning_rate": 1.597241733711008e-07,
      "loss": 0.0028,
      "step": 3200
    },
    {
      "epoch": 4.885030054384123,
      "eval_loss": 0.058813489973545074,
      "eval_runtime": 156.3015,
      "eval_samples_per_second": 7.454,
      "eval_steps_per_second": 7.454,
      "step": 3200
    },
    {
      "epoch": 4.900295773304074,
      "grad_norm": 0.22544801235198975,
      "learning_rate": 1.1998650987510295e-07,
      "loss": 0.0015,
      "step": 3210
    },
    {
      "epoch": 4.9155614922240245,
      "grad_norm": 0.014127412810921669,
      "learning_rate": 8.591726509222242e-08,
      "loss": 0.0059,
      "step": 3220
    },
    {
      "epoch": 4.930827211143975,
      "grad_norm": 0.09811969101428986,
      "learning_rate": 5.752031068830266e-08,
      "loss": 0.0016,
      "step": 3230
    },
    {
      "epoch": 4.946092930063926,
      "grad_norm": 0.08793773502111435,
      "learning_rate": 3.4798873723984605e-08,
      "loss": 0.0008,
      "step": 3240
    },
    {
      "epoch": 4.961358648983875,
      "grad_norm": 0.23520350456237793,
      "learning_rate": 1.7755536287944464e-08,
      "loss": 0.0021,
      "step": 3250
    },
    {
      "epoch": 4.961358648983875,
      "eval_loss": 0.058792199939489365,
      "eval_runtime": 156.1853,
      "eval_samples_per_second": 7.459,
      "eval_steps_per_second": 7.459,
      "step": 3250
    },
    {
      "epoch": 4.976624367903826,
      "grad_norm": 0.024279046803712845,
      "learning_rate": 6.3922352034895276e-09,
      "loss": 0.0052,
      "step": 3260
    },
    {
      "epoch": 4.991890086823776,
      "grad_norm": 0.1994946151971817,
      "learning_rate": 7.102618084620094e-10,
      "loss": 0.0023,
      "step": 3270
    },
    {
      "epoch": 4.999522946283752,
      "step": 3275,
      "total_flos": 8.22077121848279e+17,
      "train_loss": 0.04081964985958037,
      "train_runtime": 35506.1452,
      "train_samples_per_second": 1.476,
      "train_steps_per_second": 0.092
    }
  ],
  "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": 8.22077121848279e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}