{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1473,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.020366598778004074,
      "grad_norm": 7.34504383318826,
      "learning_rate": 6.756756756756758e-07,
      "loss": 0.5197,
      "step": 10
    },
    {
      "epoch": 0.04073319755600815,
      "grad_norm": 1.9502076909187314,
      "learning_rate": 1.3513513513513515e-06,
      "loss": 0.4425,
      "step": 20
    },
    {
      "epoch": 0.06109979633401222,
      "grad_norm": 1.3703382365758696,
      "learning_rate": 2.0270270270270273e-06,
      "loss": 0.3956,
      "step": 30
    },
    {
      "epoch": 0.0814663951120163,
      "grad_norm": 1.5274033164070318,
      "learning_rate": 2.702702702702703e-06,
      "loss": 0.3706,
      "step": 40
    },
    {
      "epoch": 0.10183299389002037,
      "grad_norm": 1.5387129397990396,
      "learning_rate": 3.3783783783783788e-06,
      "loss": 0.3558,
      "step": 50
    },
    {
      "epoch": 0.12219959266802444,
      "grad_norm": 1.5285386592148773,
      "learning_rate": 4.0540540540540545e-06,
      "loss": 0.3439,
      "step": 60
    },
    {
      "epoch": 0.1425661914460285,
      "grad_norm": 1.4343248038161889,
      "learning_rate": 4.72972972972973e-06,
      "loss": 0.3351,
      "step": 70
    },
    {
      "epoch": 0.1629327902240326,
      "grad_norm": 1.3758371691545166,
      "learning_rate": 4.999795773182583e-06,
      "loss": 0.3273,
      "step": 80
    },
    {
      "epoch": 0.18329938900203666,
      "grad_norm": 1.5978694880041386,
      "learning_rate": 4.998547854667808e-06,
      "loss": 0.3217,
      "step": 90
    },
    {
      "epoch": 0.20366598778004075,
      "grad_norm": 1.3851731195685373,
      "learning_rate": 4.9961661054815474e-06,
      "loss": 0.322,
      "step": 100
    },
    {
      "epoch": 0.2240325865580448,
      "grad_norm": 1.3521127973831295,
      "learning_rate": 4.992651726621268e-06,
      "loss": 0.3146,
      "step": 110
    },
    {
      "epoch": 0.24439918533604887,
      "grad_norm": 1.3516467751032153,
      "learning_rate": 4.988006490213162e-06,
      "loss": 0.3143,
      "step": 120
    },
    {
      "epoch": 0.26476578411405294,
      "grad_norm": 1.2701037565844242,
      "learning_rate": 4.9822327386185436e-06,
      "loss": 0.3077,
      "step": 130
    },
    {
      "epoch": 0.285132382892057,
      "grad_norm": 1.6979740720879781,
      "learning_rate": 4.975333383252724e-06,
      "loss": 0.307,
      "step": 140
    },
    {
      "epoch": 0.3054989816700611,
      "grad_norm": 1.3490436638938983,
      "learning_rate": 4.967311903116924e-06,
      "loss": 0.2987,
      "step": 150
    },
    {
      "epoch": 0.3258655804480652,
      "grad_norm": 1.7161350768628911,
      "learning_rate": 4.958172343043982e-06,
      "loss": 0.2999,
      "step": 160
    },
    {
      "epoch": 0.34623217922606925,
      "grad_norm": 1.382612250934794,
      "learning_rate": 4.947919311658756e-06,
      "loss": 0.2991,
      "step": 170
    },
    {
      "epoch": 0.3665987780040733,
      "grad_norm": 1.3151580723383025,
      "learning_rate": 4.936557979054219e-06,
      "loss": 0.2949,
      "step": 180
    },
    {
      "epoch": 0.3869653767820774,
      "grad_norm": 1.3980284114864507,
      "learning_rate": 4.924094074184436e-06,
      "loss": 0.2931,
      "step": 190
    },
    {
      "epoch": 0.4073319755600815,
      "grad_norm": 1.4995260622729383,
      "learning_rate": 4.910533881975742e-06,
      "loss": 0.2897,
      "step": 200
    },
    {
      "epoch": 0.42769857433808556,
      "grad_norm": 1.3091750919874185,
      "learning_rate": 4.895884240157568e-06,
      "loss": 0.2897,
      "step": 210
    },
    {
      "epoch": 0.4480651731160896,
      "grad_norm": 1.4747628264164012,
      "learning_rate": 4.880152535814507e-06,
      "loss": 0.2868,
      "step": 220
    },
    {
      "epoch": 0.4684317718940937,
      "grad_norm": 1.4470008213149266,
      "learning_rate": 4.863346701661384e-06,
      "loss": 0.2874,
      "step": 230
    },
    {
      "epoch": 0.48879837067209775,
      "grad_norm": 1.4515164854269305,
      "learning_rate": 4.845475212043178e-06,
      "loss": 0.2878,
      "step": 240
    },
    {
      "epoch": 0.5091649694501018,
      "grad_norm": 1.2578053743674917,
      "learning_rate": 4.8265470786618345e-06,
      "loss": 0.2811,
      "step": 250
    },
    {
      "epoch": 0.5295315682281059,
      "grad_norm": 1.5098983649320945,
      "learning_rate": 4.806571846032109e-06,
      "loss": 0.2773,
      "step": 260
    },
    {
      "epoch": 0.5498981670061099,
      "grad_norm": 1.260736734336587,
      "learning_rate": 4.785559586668751e-06,
      "loss": 0.28,
      "step": 270
    },
    {
      "epoch": 0.570264765784114,
      "grad_norm": 1.1734721692297405,
      "learning_rate": 4.763520896007427e-06,
      "loss": 0.2825,
      "step": 280
    },
    {
      "epoch": 0.5906313645621182,
      "grad_norm": 1.1031919455840085,
      "learning_rate": 4.74046688706198e-06,
      "loss": 0.2759,
      "step": 290
    },
    {
      "epoch": 0.6109979633401222,
      "grad_norm": 1.0944761517106407,
      "learning_rate": 4.716409184820684e-06,
      "loss": 0.2766,
      "step": 300
    },
    {
      "epoch": 0.6313645621181263,
      "grad_norm": 1.23087624035179,
      "learning_rate": 4.691359920384341e-06,
      "loss": 0.2777,
      "step": 310
    },
    {
      "epoch": 0.6517311608961304,
      "grad_norm": 1.1545647683752378,
      "learning_rate": 4.6653317248491754e-06,
      "loss": 0.2752,
      "step": 320
    },
    {
      "epoch": 0.6720977596741344,
      "grad_norm": 1.2931986965229072,
      "learning_rate": 4.638337722937599e-06,
      "loss": 0.2746,
      "step": 330
    },
    {
      "epoch": 0.6924643584521385,
      "grad_norm": 1.1273344784116983,
      "learning_rate": 4.610391526380066e-06,
      "loss": 0.2688,
      "step": 340
    },
    {
      "epoch": 0.7128309572301426,
      "grad_norm": 1.415362493265054,
      "learning_rate": 4.581507227051359e-06,
      "loss": 0.2703,
      "step": 350
    },
    {
      "epoch": 0.7331975560081466,
      "grad_norm": 1.1591318242459208,
      "learning_rate": 4.551699389864756e-06,
      "loss": 0.2665,
      "step": 360
    },
    {
      "epoch": 0.7535641547861507,
      "grad_norm": 1.0554558469893176,
      "learning_rate": 4.520983045427667e-06,
      "loss": 0.2693,
      "step": 370
    },
    {
      "epoch": 0.7739307535641547,
      "grad_norm": 1.2265072577256237,
      "learning_rate": 4.489373682462457e-06,
      "loss": 0.2663,
      "step": 380
    },
    {
      "epoch": 0.7942973523421588,
      "grad_norm": 1.2855575031097177,
      "learning_rate": 4.456887239996245e-06,
      "loss": 0.2679,
      "step": 390
    },
    {
      "epoch": 0.814663951120163,
      "grad_norm": 1.081374466408974,
      "learning_rate": 4.423540099323651e-06,
      "loss": 0.2661,
      "step": 400
    },
    {
      "epoch": 0.835030549898167,
      "grad_norm": 1.1990325052929929,
      "learning_rate": 4.3893490757465264e-06,
      "loss": 0.2648,
      "step": 410
    },
    {
      "epoch": 0.8553971486761711,
      "grad_norm": 1.2726034950554865,
      "learning_rate": 4.354331410094831e-06,
      "loss": 0.2659,
      "step": 420
    },
    {
      "epoch": 0.8757637474541752,
      "grad_norm": 1.1644285218921815,
      "learning_rate": 4.318504760032946e-06,
      "loss": 0.2613,
      "step": 430
    },
    {
      "epoch": 0.8961303462321792,
      "grad_norm": 1.236076922309445,
      "learning_rate": 4.281887191155788e-06,
      "loss": 0.2636,
      "step": 440
    },
    {
      "epoch": 0.9164969450101833,
      "grad_norm": 1.1721589750974497,
      "learning_rate": 4.244497167879234e-06,
      "loss": 0.261,
      "step": 450
    },
    {
      "epoch": 0.9368635437881874,
      "grad_norm": 1.085539973119722,
      "learning_rate": 4.206353544129437e-06,
      "loss": 0.2632,
      "step": 460
    },
    {
      "epoch": 0.9572301425661914,
      "grad_norm": 1.1393530942547674,
      "learning_rate": 4.1674755538357234e-06,
      "loss": 0.2593,
      "step": 470
    },
    {
      "epoch": 0.9775967413441955,
      "grad_norm": 1.144980396038313,
      "learning_rate": 4.12788280123189e-06,
      "loss": 0.2582,
      "step": 480
    },
    {
      "epoch": 0.9979633401221996,
      "grad_norm": 1.1413918380162131,
      "learning_rate": 4.087595250970756e-06,
      "loss": 0.2555,
      "step": 490
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.2573759853839874,
      "eval_runtime": 44.646,
      "eval_samples_per_second": 296.13,
      "eval_steps_per_second": 1.165,
      "step": 491
    },
    {
      "epoch": 1.0183299389002036,
      "grad_norm": 1.5557280500135056,
      "learning_rate": 4.0466332180569904e-06,
      "loss": 0.1936,
      "step": 500
    },
    {
      "epoch": 1.0386965376782078,
      "grad_norm": 1.3238755177486623,
      "learning_rate": 4.0050173576032644e-06,
      "loss": 0.1814,
      "step": 510
    },
    {
      "epoch": 1.0590631364562118,
      "grad_norm": 1.2066120698543659,
      "learning_rate": 3.9627686544149085e-06,
      "loss": 0.181,
      "step": 520
    },
    {
      "epoch": 1.079429735234216,
      "grad_norm": 1.088083679916152,
      "learning_rate": 3.919908412408315e-06,
      "loss": 0.1799,
      "step": 530
    },
    {
      "epoch": 1.0997963340122199,
      "grad_norm": 1.123343825207487,
      "learning_rate": 3.876458243868441e-06,
      "loss": 0.1813,
      "step": 540
    },
    {
      "epoch": 1.120162932790224,
      "grad_norm": 1.3201033277961134,
      "learning_rate": 3.832440058550809e-06,
      "loss": 0.1815,
      "step": 550
    },
    {
      "epoch": 1.140529531568228,
      "grad_norm": 1.278395529927195,
      "learning_rate": 3.7878760526335024e-06,
      "loss": 0.1781,
      "step": 560
    },
    {
      "epoch": 1.1608961303462322,
      "grad_norm": 1.11604395104044,
      "learning_rate": 3.742788697524746e-06,
      "loss": 0.1773,
      "step": 570
    },
    {
      "epoch": 1.1812627291242364,
      "grad_norm": 1.1089540254153334,
      "learning_rate": 3.6972007285316907e-06,
      "loss": 0.1806,
      "step": 580
    },
    {
      "epoch": 1.2016293279022403,
      "grad_norm": 1.1472218678631292,
      "learning_rate": 3.6511351333961297e-06,
      "loss": 0.1795,
      "step": 590
    },
    {
      "epoch": 1.2219959266802445,
      "grad_norm": 1.0597976659827861,
      "learning_rate": 3.6046151407029294e-06,
      "loss": 0.1815,
      "step": 600
    },
    {
      "epoch": 1.2423625254582484,
      "grad_norm": 1.268227483080295,
      "learning_rate": 3.557664208167006e-06,
      "loss": 0.182,
      "step": 610
    },
    {
      "epoch": 1.2627291242362526,
      "grad_norm": 1.3674823513528875,
      "learning_rate": 3.5103060108047724e-06,
      "loss": 0.181,
      "step": 620
    },
    {
      "epoch": 1.2830957230142566,
      "grad_norm": 1.2651346876174,
      "learning_rate": 3.462564428996006e-06,
      "loss": 0.1813,
      "step": 630
    },
    {
      "epoch": 1.3034623217922607,
      "grad_norm": 1.1472859968056397,
      "learning_rate": 3.414463536442161e-06,
      "loss": 0.1794,
      "step": 640
    },
    {
      "epoch": 1.3238289205702647,
      "grad_norm": 1.2499659583847704,
      "learning_rate": 3.3660275880272014e-06,
      "loss": 0.1792,
      "step": 650
    },
    {
      "epoch": 1.3441955193482689,
      "grad_norm": 1.1361447212288704,
      "learning_rate": 3.3172810075870724e-06,
      "loss": 0.1791,
      "step": 660
    },
    {
      "epoch": 1.364562118126273,
      "grad_norm": 1.0924410851905066,
      "learning_rate": 3.2682483755939735e-06,
      "loss": 0.1798,
      "step": 670
    },
    {
      "epoch": 1.384928716904277,
      "grad_norm": 1.1623338702342625,
      "learning_rate": 3.2189544167616544e-06,
      "loss": 0.1796,
      "step": 680
    },
    {
      "epoch": 1.405295315682281,
      "grad_norm": 1.3102412387590796,
      "learning_rate": 3.1694239875779708e-06,
      "loss": 0.1758,
      "step": 690
    },
    {
      "epoch": 1.4256619144602851,
      "grad_norm": 1.179561417494925,
      "learning_rate": 3.1196820637709976e-06,
      "loss": 0.1782,
      "step": 700
    },
    {
      "epoch": 1.4460285132382893,
      "grad_norm": 1.1636897041976815,
      "learning_rate": 3.0697537277150108e-06,
      "loss": 0.1796,
      "step": 710
    },
    {
      "epoch": 1.4663951120162932,
      "grad_norm": 1.0835611859399772,
      "learning_rate": 3.0196641557826993e-06,
      "loss": 0.181,
      "step": 720
    },
    {
      "epoch": 1.4867617107942974,
      "grad_norm": 1.1641341412000015,
      "learning_rate": 2.969438605649973e-06,
      "loss": 0.1784,
      "step": 730
    },
    {
      "epoch": 1.5071283095723014,
      "grad_norm": 1.4309684703117318,
      "learning_rate": 2.9191024035597715e-06,
      "loss": 0.1785,
      "step": 740
    },
    {
      "epoch": 1.5274949083503055,
      "grad_norm": 1.1211318477711534,
      "learning_rate": 2.868680931551307e-06,
      "loss": 0.1785,
      "step": 750
    },
    {
      "epoch": 1.5478615071283097,
      "grad_norm": 1.0844733637132042,
      "learning_rate": 2.818199614661163e-06,
      "loss": 0.1784,
      "step": 760
    },
    {
      "epoch": 1.5682281059063137,
      "grad_norm": 1.0733420166347605,
      "learning_rate": 2.7676839081027153e-06,
      "loss": 0.177,
      "step": 770
    },
    {
      "epoch": 1.5885947046843176,
      "grad_norm": 1.0406238069755496,
      "learning_rate": 2.7171592844303377e-06,
      "loss": 0.177,
      "step": 780
    },
    {
      "epoch": 1.6089613034623218,
      "grad_norm": 1.0173564043856271,
      "learning_rate": 2.6666512206948574e-06,
      "loss": 0.1776,
      "step": 790
    },
    {
      "epoch": 1.629327902240326,
      "grad_norm": 1.098211071807489,
      "learning_rate": 2.6161851855967546e-06,
      "loss": 0.1766,
      "step": 800
    },
    {
      "epoch": 1.64969450101833,
      "grad_norm": 1.23874084216799,
      "learning_rate": 2.565786626643557e-06,
      "loss": 0.1805,
      "step": 810
    },
    {
      "epoch": 1.6700610997963339,
      "grad_norm": 1.0816973984606886,
      "learning_rate": 2.5154809573179374e-06,
      "loss": 0.1797,
      "step": 820
    },
    {
      "epoch": 1.690427698574338,
      "grad_norm": 1.0857968496669568,
      "learning_rate": 2.465293544262953e-06,
      "loss": 0.1767,
      "step": 830
    },
    {
      "epoch": 1.7107942973523422,
      "grad_norm": 1.0614368562971397,
      "learning_rate": 2.4152496944909185e-06,
      "loss": 0.1747,
      "step": 840
    },
    {
      "epoch": 1.7311608961303462,
      "grad_norm": 1.1079176704743572,
      "learning_rate": 2.365374642622334e-06,
      "loss": 0.1749,
      "step": 850
    },
    {
      "epoch": 1.7515274949083504,
      "grad_norm": 1.149185145242321,
      "learning_rate": 2.315693538161327e-06,
      "loss": 0.1746,
      "step": 860
    },
    {
      "epoch": 1.7718940936863543,
      "grad_norm": 1.1364076771423424,
      "learning_rate": 2.2662314328140077e-06,
      "loss": 0.1757,
      "step": 870
    },
    {
      "epoch": 1.7922606924643585,
      "grad_norm": 1.0481149270348264,
      "learning_rate": 2.217013267856143e-06,
      "loss": 0.175,
      "step": 880
    },
    {
      "epoch": 1.8126272912423627,
      "grad_norm": 1.121398301379226,
      "learning_rate": 2.168063861556515e-06,
      "loss": 0.1739,
      "step": 890
    },
    {
      "epoch": 1.8329938900203666,
      "grad_norm": 1.1627216220108434,
      "learning_rate": 2.1194078966623004e-06,
      "loss": 0.1739,
      "step": 900
    },
    {
      "epoch": 1.8533604887983706,
      "grad_norm": 1.0423836344485042,
      "learning_rate": 2.0710699079527936e-06,
      "loss": 0.1737,
      "step": 910
    },
    {
      "epoch": 1.8737270875763747,
      "grad_norm": 1.1321126420156151,
      "learning_rate": 2.0230742698677407e-06,
      "loss": 0.1752,
      "step": 920
    },
    {
      "epoch": 1.894093686354379,
      "grad_norm": 1.0982860659977216,
      "learning_rate": 1.9754451842165187e-06,
      "loss": 0.1732,
      "step": 930
    },
    {
      "epoch": 1.9144602851323829,
      "grad_norm": 1.0607568085093442,
      "learning_rate": 1.9282066679743756e-06,
      "loss": 0.1772,
      "step": 940
    },
    {
      "epoch": 1.9348268839103868,
      "grad_norm": 1.1195338613324972,
      "learning_rate": 1.8813825411718608e-06,
      "loss": 0.1739,
      "step": 950
    },
    {
      "epoch": 1.955193482688391,
      "grad_norm": 1.2105165500055401,
      "learning_rate": 1.834996414883573e-06,
      "loss": 0.1745,
      "step": 960
    },
    {
      "epoch": 1.9755600814663952,
      "grad_norm": 1.0195074211250463,
      "learning_rate": 1.789071679322268e-06,
      "loss": 0.1726,
      "step": 970
    },
    {
      "epoch": 1.9959266802443993,
      "grad_norm": 1.1218629665522688,
      "learning_rate": 1.7436314920443475e-06,
      "loss": 0.1732,
      "step": 980
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.2461756318807602,
      "eval_runtime": 44.0297,
      "eval_samples_per_second": 300.275,
      "eval_steps_per_second": 1.181,
      "step": 982
    },
    {
      "epoch": 2.016293279022403,
      "grad_norm": 1.3922840302017268,
      "learning_rate": 1.6986987662726467e-06,
      "loss": 0.1202,
      "step": 990
    },
    {
      "epoch": 2.0366598778004072,
      "grad_norm": 1.2614971724968627,
      "learning_rate": 1.6542961593424448e-06,
      "loss": 0.1048,
      "step": 1000
    },
    {
      "epoch": 2.0570264765784114,
      "grad_norm": 1.1248418232642112,
      "learning_rate": 1.6104460612764963e-06,
      "loss": 0.1039,
      "step": 1010
    },
    {
      "epoch": 2.0773930753564156,
      "grad_norm": 1.1221931057869212,
      "learning_rate": 1.5671705834948596e-06,
      "loss": 0.1052,
      "step": 1020
    },
    {
      "epoch": 2.0977596741344193,
      "grad_norm": 1.3233889385769144,
      "learning_rate": 1.5244915476652073e-06,
      "loss": 0.1045,
      "step": 1030
    },
    {
      "epoch": 2.1181262729124235,
      "grad_norm": 1.219740632707147,
      "learning_rate": 1.4824304746992488e-06,
      "loss": 0.1045,
      "step": 1040
    },
    {
      "epoch": 2.1384928716904277,
      "grad_norm": 1.1317738274870492,
      "learning_rate": 1.441008573900804e-06,
      "loss": 0.1038,
      "step": 1050
    },
    {
      "epoch": 2.158859470468432,
      "grad_norm": 1.1214324601378942,
      "learning_rate": 1.4002467322710097e-06,
      "loss": 0.1041,
      "step": 1060
    },
    {
      "epoch": 2.179226069246436,
      "grad_norm": 1.113920354990809,
      "learning_rate": 1.3601655039760452e-06,
      "loss": 0.1051,
      "step": 1070
    },
    {
      "epoch": 2.1995926680244398,
      "grad_norm": 1.1594270616003397,
      "learning_rate": 1.3207850999826893e-06,
      "loss": 0.1048,
      "step": 1080
    },
    {
      "epoch": 2.219959266802444,
      "grad_norm": 1.026595442656351,
      "learning_rate": 1.2821253778669374e-06,
      "loss": 0.104,
      "step": 1090
    },
    {
      "epoch": 2.240325865580448,
      "grad_norm": 1.0868748314482166,
      "learning_rate": 1.2442058318008143e-06,
      "loss": 0.1045,
      "step": 1100
    },
    {
      "epoch": 2.2606924643584523,
      "grad_norm": 1.1163655110568669,
      "learning_rate": 1.2070455827224361e-06,
      "loss": 0.1028,
      "step": 1110
    },
    {
      "epoch": 2.281059063136456,
      "grad_norm": 1.1265872495516769,
      "learning_rate": 1.170663368694271e-06,
      "loss": 0.1049,
      "step": 1120
    },
    {
      "epoch": 2.30142566191446,
      "grad_norm": 1.0787109584417223,
      "learning_rate": 1.1350775354544707e-06,
      "loss": 0.104,
      "step": 1130
    },
    {
      "epoch": 2.3217922606924644,
      "grad_norm": 1.2647428474671398,
      "learning_rate": 1.1003060271660259e-06,
      "loss": 0.104,
      "step": 1140
    },
    {
      "epoch": 2.3421588594704685,
      "grad_norm": 1.1925610582272053,
      "learning_rate": 1.06636637736842e-06,
      "loss": 0.1042,
      "step": 1150
    },
    {
      "epoch": 2.3625254582484727,
      "grad_norm": 1.0856982970751745,
      "learning_rate": 1.0332757001363402e-06,
      "loss": 0.1043,
      "step": 1160
    },
    {
      "epoch": 2.3828920570264764,
      "grad_norm": 1.1357554208144642,
      "learning_rate": 1.0010506814498994e-06,
      "loss": 0.1031,
      "step": 1170
    },
    {
      "epoch": 2.4032586558044806,
      "grad_norm": 1.185189241987044,
      "learning_rate": 9.697075707807282e-07,
      "loss": 0.1041,
      "step": 1180
    },
    {
      "epoch": 2.423625254582485,
      "grad_norm": 1.134389319376477,
      "learning_rate": 9.392621728981774e-07,
      "loss": 0.1045,
      "step": 1190
    },
    {
      "epoch": 2.443991853360489,
      "grad_norm": 1.1120088243562505,
      "learning_rate": 9.097298398997584e-07,
      "loss": 0.1027,
      "step": 1200
    },
    {
      "epoch": 2.4643584521384927,
      "grad_norm": 1.1115040698718408,
      "learning_rate": 8.811254634698468e-07,
      "loss": 0.105,
      "step": 1210
    },
    {
      "epoch": 2.484725050916497,
      "grad_norm": 1.0851503339063027,
      "learning_rate": 8.534634673705461e-07,
      "loss": 0.1038,
      "step": 1220
    },
    {
      "epoch": 2.505091649694501,
      "grad_norm": 1.1183787883183298,
      "learning_rate": 8.267578001685048e-07,
      "loss": 0.102,
      "step": 1230
    },
    {
      "epoch": 2.525458248472505,
      "grad_norm": 1.1421505006564803,
      "learning_rate": 8.010219282013471e-07,
      "loss": 0.1047,
      "step": 1240
    },
    {
      "epoch": 2.5458248472505094,
      "grad_norm": 1.1411521931249191,
      "learning_rate": 7.76268828787271e-07,
      "loss": 0.1037,
      "step": 1250
    },
    {
      "epoch": 2.566191446028513,
      "grad_norm": 1.0934492001333824,
      "learning_rate": 7.525109836812336e-07,
      "loss": 0.1028,
      "step": 1260
    },
    {
      "epoch": 2.5865580448065173,
      "grad_norm": 1.098352714811253,
      "learning_rate": 7.297603727810222e-07,
      "loss": 0.1034,
      "step": 1270
    },
    {
      "epoch": 2.6069246435845215,
      "grad_norm": 1.1012241466696564,
      "learning_rate": 7.0802846808639e-07,
      "loss": 0.1039,
      "step": 1280
    },
    {
      "epoch": 2.627291242362525,
      "grad_norm": 1.082718927412021,
      "learning_rate": 6.873262279142977e-07,
      "loss": 0.1026,
      "step": 1290
    },
    {
      "epoch": 2.6476578411405294,
      "grad_norm": 1.0913895786684868,
      "learning_rate": 6.676640913731799e-07,
      "loss": 0.1026,
      "step": 1300
    },
    {
      "epoch": 2.6680244399185336,
      "grad_norm": 1.114449983311916,
      "learning_rate": 6.490519730990235e-07,
      "loss": 0.1036,
      "step": 1310
    },
    {
      "epoch": 2.6883910386965377,
      "grad_norm": 1.0981797148792212,
      "learning_rate": 6.314992582559093e-07,
      "loss": 0.1044,
      "step": 1320
    },
    {
      "epoch": 2.708757637474542,
      "grad_norm": 1.1031210331032415,
      "learning_rate": 6.150147978035423e-07,
      "loss": 0.1028,
      "step": 1330
    },
    {
      "epoch": 2.729124236252546,
      "grad_norm": 1.0231804838279508,
      "learning_rate": 5.996069040341544e-07,
      "loss": 0.1007,
      "step": 1340
    },
    {
      "epoch": 2.74949083503055,
      "grad_norm": 1.1134674052519227,
      "learning_rate": 5.852833463810277e-07,
      "loss": 0.1032,
      "step": 1350
    },
    {
      "epoch": 2.769857433808554,
      "grad_norm": 1.0615744074260045,
      "learning_rate": 5.720513475007569e-07,
      "loss": 0.1042,
      "step": 1360
    },
    {
      "epoch": 2.790224032586558,
      "grad_norm": 1.0551133440758214,
      "learning_rate": 5.599175796312243e-07,
      "loss": 0.1037,
      "step": 1370
    },
    {
      "epoch": 2.810590631364562,
      "grad_norm": 1.085953698001666,
      "learning_rate": 5.488881612271214e-07,
      "loss": 0.1027,
      "step": 1380
    },
    {
      "epoch": 2.830957230142566,
      "grad_norm": 1.069753873055028,
      "learning_rate": 5.389686538747183e-07,
      "loss": 0.1036,
      "step": 1390
    },
    {
      "epoch": 2.8513238289205702,
      "grad_norm": 1.1482168307825666,
      "learning_rate": 5.301640594874314e-07,
      "loss": 0.1035,
      "step": 1400
    },
    {
      "epoch": 2.8716904276985744,
      "grad_norm": 1.0724177471486462,
      "learning_rate": 5.224788177836086e-07,
      "loss": 0.1028,
      "step": 1410
    },
    {
      "epoch": 2.8920570264765786,
      "grad_norm": 1.0644407890849343,
      "learning_rate": 5.159168040478003e-07,
      "loss": 0.1036,
      "step": 1420
    },
    {
      "epoch": 2.9124236252545828,
      "grad_norm": 1.1122712461112416,
      "learning_rate": 5.104813271766455e-07,
      "loss": 0.1011,
      "step": 1430
    },
    {
      "epoch": 2.9327902240325865,
      "grad_norm": 1.0864458509689723,
      "learning_rate": 5.061751280103621e-07,
      "loss": 0.1027,
      "step": 1440
    },
    {
      "epoch": 2.9531568228105907,
      "grad_norm": 1.0960544633048377,
      "learning_rate": 5.030003779506757e-07,
      "loss": 0.1016,
      "step": 1450
    },
    {
      "epoch": 2.973523421588595,
      "grad_norm": 1.0723267354322656,
      "learning_rate": 5.009586778658912e-07,
      "loss": 0.1012,
      "step": 1460
    },
    {
      "epoch": 2.9938900203665986,
      "grad_norm": 1.0958889614425518,
      "learning_rate": 5.000510572836534e-07,
      "loss": 0.1013,
      "step": 1470
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.2728930115699768,
      "eval_runtime": 45.1101,
      "eval_samples_per_second": 293.083,
      "eval_steps_per_second": 1.153,
      "step": 1473
    },
    {
      "epoch": 3.0,
      "step": 1473,
      "total_flos": 2467329862533120.0,
      "train_loss": 0.19335093406374892,
      "train_runtime": 8782.5965,
      "train_samples_per_second": 85.801,
      "train_steps_per_second": 0.168
    }
  ],
  "logging_steps": 10,
  "max_steps": 1473,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2467329862533120.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}