{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.6253908692933083,
  "eval_steps": 500,
  "global_step": 16000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003908692933083177,
      "grad_norm": 0.47716134786605835,
      "learning_rate": 4.98240400406663e-05,
      "loss": 1.3868,
      "step": 100
    },
    {
      "epoch": 0.007817385866166354,
      "grad_norm": 0.28900113701820374,
      "learning_rate": 4.9628528974739973e-05,
      "loss": 1.3348,
      "step": 200
    },
    {
      "epoch": 0.01172607879924953,
      "grad_norm": 0.37652796506881714,
      "learning_rate": 4.943301790881364e-05,
      "loss": 1.3298,
      "step": 300
    },
    {
      "epoch": 0.015634771732332707,
      "grad_norm": 0.24992702901363373,
      "learning_rate": 4.923750684288731e-05,
      "loss": 1.3309,
      "step": 400
    },
    {
      "epoch": 0.019543464665415886,
      "grad_norm": 0.3237004280090332,
      "learning_rate": 4.904199577696098e-05,
      "loss": 1.3228,
      "step": 500
    },
    {
      "epoch": 0.02345215759849906,
      "grad_norm": 0.259011447429657,
      "learning_rate": 4.884648471103465e-05,
      "loss": 1.3166,
      "step": 600
    },
    {
      "epoch": 0.02736085053158224,
      "grad_norm": 0.34121841192245483,
      "learning_rate": 4.8650973645108314e-05,
      "loss": 1.3145,
      "step": 700
    },
    {
      "epoch": 0.031269543464665414,
      "grad_norm": 0.41398781538009644,
      "learning_rate": 4.845546257918198e-05,
      "loss": 1.3081,
      "step": 800
    },
    {
      "epoch": 0.03517823639774859,
      "grad_norm": 0.4289904832839966,
      "learning_rate": 4.825995151325565e-05,
      "loss": 1.3073,
      "step": 900
    },
    {
      "epoch": 0.03908692933083177,
      "grad_norm": 0.21226634085178375,
      "learning_rate": 4.8064440447329324e-05,
      "loss": 1.3057,
      "step": 1000
    },
    {
      "epoch": 0.04299562226391495,
      "grad_norm": 0.2929254174232483,
      "learning_rate": 4.786892938140299e-05,
      "loss": 1.3034,
      "step": 1100
    },
    {
      "epoch": 0.04690431519699812,
      "grad_norm": 0.22025109827518463,
      "learning_rate": 4.767341831547666e-05,
      "loss": 1.3048,
      "step": 1200
    },
    {
      "epoch": 0.0508130081300813,
      "grad_norm": 0.36455097794532776,
      "learning_rate": 4.747790724955033e-05,
      "loss": 1.3028,
      "step": 1300
    },
    {
      "epoch": 0.05472170106316448,
      "grad_norm": 0.24527208507061005,
      "learning_rate": 4.728239618362399e-05,
      "loss": 1.2996,
      "step": 1400
    },
    {
      "epoch": 0.05863039399624766,
      "grad_norm": 0.33187615871429443,
      "learning_rate": 4.7086885117697664e-05,
      "loss": 1.3001,
      "step": 1500
    },
    {
      "epoch": 0.06253908692933083,
      "grad_norm": 0.502117931842804,
      "learning_rate": 4.689137405177133e-05,
      "loss": 1.2954,
      "step": 1600
    },
    {
      "epoch": 0.06644777986241401,
      "grad_norm": 0.289112389087677,
      "learning_rate": 4.6695862985845e-05,
      "loss": 1.2969,
      "step": 1700
    },
    {
      "epoch": 0.07035647279549719,
      "grad_norm": 0.24350133538246155,
      "learning_rate": 4.6500351919918674e-05,
      "loss": 1.2929,
      "step": 1800
    },
    {
      "epoch": 0.07426516572858036,
      "grad_norm": 0.2194579690694809,
      "learning_rate": 4.630484085399234e-05,
      "loss": 1.2913,
      "step": 1900
    },
    {
      "epoch": 0.07817385866166354,
      "grad_norm": 0.3578532636165619,
      "learning_rate": 4.6109329788066005e-05,
      "loss": 1.2946,
      "step": 2000
    },
    {
      "epoch": 0.08208255159474671,
      "grad_norm": 0.22988037765026093,
      "learning_rate": 4.591381872213967e-05,
      "loss": 1.2906,
      "step": 2100
    },
    {
      "epoch": 0.0859912445278299,
      "grad_norm": 0.44038262963294983,
      "learning_rate": 4.571830765621334e-05,
      "loss": 1.2941,
      "step": 2200
    },
    {
      "epoch": 0.08989993746091307,
      "grad_norm": 0.27333030104637146,
      "learning_rate": 4.5522796590287015e-05,
      "loss": 1.2903,
      "step": 2300
    },
    {
      "epoch": 0.09380863039399624,
      "grad_norm": 0.23196421563625336,
      "learning_rate": 4.532728552436068e-05,
      "loss": 1.2897,
      "step": 2400
    },
    {
      "epoch": 0.09771732332707943,
      "grad_norm": 0.25054216384887695,
      "learning_rate": 4.513177445843435e-05,
      "loss": 1.2891,
      "step": 2500
    },
    {
      "epoch": 0.1016260162601626,
      "grad_norm": 0.17129705846309662,
      "learning_rate": 4.493626339250802e-05,
      "loss": 1.2874,
      "step": 2600
    },
    {
      "epoch": 0.10553470919324578,
      "grad_norm": 0.5260087251663208,
      "learning_rate": 4.474075232658168e-05,
      "loss": 1.289,
      "step": 2700
    },
    {
      "epoch": 0.10944340212632896,
      "grad_norm": 0.2251499593257904,
      "learning_rate": 4.4545241260655355e-05,
      "loss": 1.2882,
      "step": 2800
    },
    {
      "epoch": 0.11335209505941213,
      "grad_norm": 0.24193763732910156,
      "learning_rate": 4.434973019472902e-05,
      "loss": 1.2886,
      "step": 2900
    },
    {
      "epoch": 0.11726078799249531,
      "grad_norm": 0.2361423224210739,
      "learning_rate": 4.415421912880269e-05,
      "loss": 1.2863,
      "step": 3000
    },
    {
      "epoch": 0.12116948092557848,
      "grad_norm": 0.22438161075115204,
      "learning_rate": 4.3958708062876365e-05,
      "loss": 1.2861,
      "step": 3100
    },
    {
      "epoch": 0.12507817385866166,
      "grad_norm": 0.30306315422058105,
      "learning_rate": 4.376319699695003e-05,
      "loss": 1.2838,
      "step": 3200
    },
    {
      "epoch": 0.12898686679174484,
      "grad_norm": 0.2526134252548218,
      "learning_rate": 4.35676859310237e-05,
      "loss": 1.288,
      "step": 3300
    },
    {
      "epoch": 0.13289555972482803,
      "grad_norm": 0.27540501952171326,
      "learning_rate": 4.337217486509736e-05,
      "loss": 1.2878,
      "step": 3400
    },
    {
      "epoch": 0.13680425265791119,
      "grad_norm": 0.2769360840320587,
      "learning_rate": 4.3176663799171034e-05,
      "loss": 1.2844,
      "step": 3500
    },
    {
      "epoch": 0.14071294559099437,
      "grad_norm": 0.3868384063243866,
      "learning_rate": 4.2981152733244706e-05,
      "loss": 1.2856,
      "step": 3600
    },
    {
      "epoch": 0.14462163852407756,
      "grad_norm": 0.2426549792289734,
      "learning_rate": 4.278564166731837e-05,
      "loss": 1.2843,
      "step": 3700
    },
    {
      "epoch": 0.1485303314571607,
      "grad_norm": 0.28486019372940063,
      "learning_rate": 4.2590130601392044e-05,
      "loss": 1.2812,
      "step": 3800
    },
    {
      "epoch": 0.1524390243902439,
      "grad_norm": 0.3195390999317169,
      "learning_rate": 4.239461953546571e-05,
      "loss": 1.2845,
      "step": 3900
    },
    {
      "epoch": 0.15634771732332708,
      "grad_norm": 0.3183465003967285,
      "learning_rate": 4.219910846953938e-05,
      "loss": 1.281,
      "step": 4000
    },
    {
      "epoch": 0.16025641025641027,
      "grad_norm": 0.30105575919151306,
      "learning_rate": 4.2003597403613047e-05,
      "loss": 1.2817,
      "step": 4100
    },
    {
      "epoch": 0.16416510318949343,
      "grad_norm": 0.3548736274242401,
      "learning_rate": 4.180808633768671e-05,
      "loss": 1.2805,
      "step": 4200
    },
    {
      "epoch": 0.1680737961225766,
      "grad_norm": 0.21018876135349274,
      "learning_rate": 4.1612575271760384e-05,
      "loss": 1.28,
      "step": 4300
    },
    {
      "epoch": 0.1719824890556598,
      "grad_norm": 0.2551437318325043,
      "learning_rate": 4.141706420583405e-05,
      "loss": 1.279,
      "step": 4400
    },
    {
      "epoch": 0.17589118198874296,
      "grad_norm": 0.21301493048667908,
      "learning_rate": 4.122155313990772e-05,
      "loss": 1.2802,
      "step": 4500
    },
    {
      "epoch": 0.17979987492182614,
      "grad_norm": 0.3330891728401184,
      "learning_rate": 4.1026042073981394e-05,
      "loss": 1.2784,
      "step": 4600
    },
    {
      "epoch": 0.18370856785490933,
      "grad_norm": 0.20550762116909027,
      "learning_rate": 4.083053100805506e-05,
      "loss": 1.2802,
      "step": 4700
    },
    {
      "epoch": 0.18761726078799248,
      "grad_norm": 0.2603992521762848,
      "learning_rate": 4.0635019942128725e-05,
      "loss": 1.2792,
      "step": 4800
    },
    {
      "epoch": 0.19152595372107567,
      "grad_norm": 0.17917224764823914,
      "learning_rate": 4.04395088762024e-05,
      "loss": 1.2803,
      "step": 4900
    },
    {
      "epoch": 0.19543464665415886,
      "grad_norm": 0.17226773500442505,
      "learning_rate": 4.024399781027606e-05,
      "loss": 1.2775,
      "step": 5000
    },
    {
      "epoch": 0.199343339587242,
      "grad_norm": 0.28991296887397766,
      "learning_rate": 4.0048486744349735e-05,
      "loss": 1.2801,
      "step": 5100
    },
    {
      "epoch": 0.2032520325203252,
      "grad_norm": 0.2766139805316925,
      "learning_rate": 3.98529756784234e-05,
      "loss": 1.2773,
      "step": 5200
    },
    {
      "epoch": 0.20716072545340838,
      "grad_norm": 0.3101244866847992,
      "learning_rate": 3.965746461249707e-05,
      "loss": 1.2786,
      "step": 5300
    },
    {
      "epoch": 0.21106941838649157,
      "grad_norm": 0.2537230849266052,
      "learning_rate": 3.946195354657074e-05,
      "loss": 1.2782,
      "step": 5400
    },
    {
      "epoch": 0.21497811131957473,
      "grad_norm": 0.388886958360672,
      "learning_rate": 3.92664424806444e-05,
      "loss": 1.2766,
      "step": 5500
    },
    {
      "epoch": 0.2188868042526579,
      "grad_norm": 0.3374210596084595,
      "learning_rate": 3.9070931414718075e-05,
      "loss": 1.273,
      "step": 5600
    },
    {
      "epoch": 0.2227954971857411,
      "grad_norm": 0.34907710552215576,
      "learning_rate": 3.887542034879174e-05,
      "loss": 1.2772,
      "step": 5700
    },
    {
      "epoch": 0.22670419011882426,
      "grad_norm": 0.22610360383987427,
      "learning_rate": 3.867990928286541e-05,
      "loss": 1.2744,
      "step": 5800
    },
    {
      "epoch": 0.23061288305190744,
      "grad_norm": 0.27353575825691223,
      "learning_rate": 3.8484398216939085e-05,
      "loss": 1.2755,
      "step": 5900
    },
    {
      "epoch": 0.23452157598499063,
      "grad_norm": 0.3231126368045807,
      "learning_rate": 3.828888715101275e-05,
      "loss": 1.2746,
      "step": 6000
    },
    {
      "epoch": 0.23843026891807378,
      "grad_norm": 0.3560299575328827,
      "learning_rate": 3.8093376085086416e-05,
      "loss": 1.2725,
      "step": 6100
    },
    {
      "epoch": 0.24233896185115697,
      "grad_norm": 0.2950970530509949,
      "learning_rate": 3.789786501916008e-05,
      "loss": 1.27,
      "step": 6200
    },
    {
      "epoch": 0.24624765478424016,
      "grad_norm": 0.25703492760658264,
      "learning_rate": 3.770235395323375e-05,
      "loss": 1.2705,
      "step": 6300
    },
    {
      "epoch": 0.2501563477173233,
      "grad_norm": 0.4161667227745056,
      "learning_rate": 3.7506842887307426e-05,
      "loss": 1.2725,
      "step": 6400
    },
    {
      "epoch": 0.2540650406504065,
      "grad_norm": 0.4292469322681427,
      "learning_rate": 3.731133182138109e-05,
      "loss": 1.2736,
      "step": 6500
    },
    {
      "epoch": 0.2579737335834897,
      "grad_norm": 0.28723031282424927,
      "learning_rate": 3.711582075545476e-05,
      "loss": 1.2696,
      "step": 6600
    },
    {
      "epoch": 0.26188242651657284,
      "grad_norm": 0.3981460928916931,
      "learning_rate": 3.692030968952843e-05,
      "loss": 1.2712,
      "step": 6700
    },
    {
      "epoch": 0.26579111944965605,
      "grad_norm": 0.509950578212738,
      "learning_rate": 3.6724798623602094e-05,
      "loss": 1.272,
      "step": 6800
    },
    {
      "epoch": 0.2696998123827392,
      "grad_norm": 0.2820459008216858,
      "learning_rate": 3.6529287557675766e-05,
      "loss": 1.2704,
      "step": 6900
    },
    {
      "epoch": 0.27360850531582237,
      "grad_norm": 0.2927381098270416,
      "learning_rate": 3.633377649174943e-05,
      "loss": 1.268,
      "step": 7000
    },
    {
      "epoch": 0.2775171982489056,
      "grad_norm": 0.2999459505081177,
      "learning_rate": 3.6138265425823104e-05,
      "loss": 1.2673,
      "step": 7100
    },
    {
      "epoch": 0.28142589118198874,
      "grad_norm": 0.3287661373615265,
      "learning_rate": 3.5942754359896776e-05,
      "loss": 1.2693,
      "step": 7200
    },
    {
      "epoch": 0.2853345841150719,
      "grad_norm": 0.4145341217517853,
      "learning_rate": 3.574724329397044e-05,
      "loss": 1.2655,
      "step": 7300
    },
    {
      "epoch": 0.2892432770481551,
      "grad_norm": 0.41897234320640564,
      "learning_rate": 3.555173222804411e-05,
      "loss": 1.2633,
      "step": 7400
    },
    {
      "epoch": 0.29315196998123827,
      "grad_norm": 0.6244714856147766,
      "learning_rate": 3.535622116211777e-05,
      "loss": 1.2659,
      "step": 7500
    },
    {
      "epoch": 0.2970606629143214,
      "grad_norm": 0.41341808438301086,
      "learning_rate": 3.5160710096191444e-05,
      "loss": 1.2635,
      "step": 7600
    },
    {
      "epoch": 0.30096935584740464,
      "grad_norm": 0.39480525255203247,
      "learning_rate": 3.4965199030265117e-05,
      "loss": 1.2633,
      "step": 7700
    },
    {
      "epoch": 0.3048780487804878,
      "grad_norm": 0.5332251787185669,
      "learning_rate": 3.476968796433878e-05,
      "loss": 1.262,
      "step": 7800
    },
    {
      "epoch": 0.30878674171357096,
      "grad_norm": 0.46669408679008484,
      "learning_rate": 3.4574176898412454e-05,
      "loss": 1.2587,
      "step": 7900
    },
    {
      "epoch": 0.31269543464665417,
      "grad_norm": 0.46353575587272644,
      "learning_rate": 3.437866583248612e-05,
      "loss": 1.2611,
      "step": 8000
    },
    {
      "epoch": 0.3166041275797373,
      "grad_norm": 0.5144259333610535,
      "learning_rate": 3.418315476655979e-05,
      "loss": 1.2603,
      "step": 8100
    },
    {
      "epoch": 0.32051282051282054,
      "grad_norm": 0.501035749912262,
      "learning_rate": 3.398764370063346e-05,
      "loss": 1.26,
      "step": 8200
    },
    {
      "epoch": 0.3244215134459037,
      "grad_norm": 0.6831699013710022,
      "learning_rate": 3.379213263470712e-05,
      "loss": 1.2598,
      "step": 8300
    },
    {
      "epoch": 0.32833020637898686,
      "grad_norm": 0.5810164213180542,
      "learning_rate": 3.3596621568780795e-05,
      "loss": 1.2561,
      "step": 8400
    },
    {
      "epoch": 0.33223889931207007,
      "grad_norm": 0.45409688353538513,
      "learning_rate": 3.340111050285447e-05,
      "loss": 1.2552,
      "step": 8500
    },
    {
      "epoch": 0.3361475922451532,
      "grad_norm": 0.6193764805793762,
      "learning_rate": 3.320559943692813e-05,
      "loss": 1.2538,
      "step": 8600
    },
    {
      "epoch": 0.3400562851782364,
      "grad_norm": 0.4400370717048645,
      "learning_rate": 3.3010088371001805e-05,
      "loss": 1.2548,
      "step": 8700
    },
    {
      "epoch": 0.3439649781113196,
      "grad_norm": 0.7784037590026855,
      "learning_rate": 3.281457730507547e-05,
      "loss": 1.2537,
      "step": 8800
    },
    {
      "epoch": 0.34787367104440275,
      "grad_norm": 0.7589293718338013,
      "learning_rate": 3.2619066239149135e-05,
      "loss": 1.2519,
      "step": 8900
    },
    {
      "epoch": 0.3517823639774859,
      "grad_norm": 0.4971660375595093,
      "learning_rate": 3.242355517322281e-05,
      "loss": 1.25,
      "step": 9000
    },
    {
      "epoch": 0.3556910569105691,
      "grad_norm": 0.5253425240516663,
      "learning_rate": 3.222804410729647e-05,
      "loss": 1.2497,
      "step": 9100
    },
    {
      "epoch": 0.3595997498436523,
      "grad_norm": 0.7438779473304749,
      "learning_rate": 3.2032533041370145e-05,
      "loss": 1.2458,
      "step": 9200
    },
    {
      "epoch": 0.36350844277673544,
      "grad_norm": 0.9425673484802246,
      "learning_rate": 3.183702197544381e-05,
      "loss": 1.2422,
      "step": 9300
    },
    {
      "epoch": 0.36741713570981865,
      "grad_norm": 0.5804303288459778,
      "learning_rate": 3.164151090951748e-05,
      "loss": 1.2445,
      "step": 9400
    },
    {
      "epoch": 0.3713258286429018,
      "grad_norm": 0.7079547047615051,
      "learning_rate": 3.144599984359115e-05,
      "loss": 1.2402,
      "step": 9500
    },
    {
      "epoch": 0.37523452157598497,
      "grad_norm": 1.088934063911438,
      "learning_rate": 3.1250488777664814e-05,
      "loss": 1.2453,
      "step": 9600
    },
    {
      "epoch": 0.3791432145090682,
      "grad_norm": 0.48442408442497253,
      "learning_rate": 3.1054977711738486e-05,
      "loss": 1.2469,
      "step": 9700
    },
    {
      "epoch": 0.38305190744215134,
      "grad_norm": 0.803571879863739,
      "learning_rate": 3.085946664581215e-05,
      "loss": 1.2448,
      "step": 9800
    },
    {
      "epoch": 0.3869606003752345,
      "grad_norm": 0.7402758598327637,
      "learning_rate": 3.0663955579885823e-05,
      "loss": 1.239,
      "step": 9900
    },
    {
      "epoch": 0.3908692933083177,
      "grad_norm": 0.874411404132843,
      "learning_rate": 3.0468444513959492e-05,
      "loss": 1.2424,
      "step": 10000
    },
    {
      "epoch": 0.39477798624140087,
      "grad_norm": 0.8369914293289185,
      "learning_rate": 3.0272933448033158e-05,
      "loss": 1.2406,
      "step": 10100
    },
    {
      "epoch": 0.398686679174484,
      "grad_norm": 0.9173115491867065,
      "learning_rate": 3.007742238210683e-05,
      "loss": 1.2338,
      "step": 10200
    },
    {
      "epoch": 0.40259537210756724,
      "grad_norm": 1.1409565210342407,
      "learning_rate": 2.98819113161805e-05,
      "loss": 1.2332,
      "step": 10300
    },
    {
      "epoch": 0.4065040650406504,
      "grad_norm": 1.2753782272338867,
      "learning_rate": 2.9686400250254164e-05,
      "loss": 1.23,
      "step": 10400
    },
    {
      "epoch": 0.4104127579737336,
      "grad_norm": 1.3038467168807983,
      "learning_rate": 2.9490889184327836e-05,
      "loss": 1.229,
      "step": 10500
    },
    {
      "epoch": 0.41432145090681677,
      "grad_norm": 0.7983734607696533,
      "learning_rate": 2.92953781184015e-05,
      "loss": 1.2318,
      "step": 10600
    },
    {
      "epoch": 0.4182301438398999,
      "grad_norm": 0.8295246362686157,
      "learning_rate": 2.909986705247517e-05,
      "loss": 1.2335,
      "step": 10700
    },
    {
      "epoch": 0.42213883677298314,
      "grad_norm": 0.8072339296340942,
      "learning_rate": 2.8904355986548843e-05,
      "loss": 1.2241,
      "step": 10800
    },
    {
      "epoch": 0.4260475297060663,
      "grad_norm": 0.7493770122528076,
      "learning_rate": 2.8708844920622508e-05,
      "loss": 1.2269,
      "step": 10900
    },
    {
      "epoch": 0.42995622263914945,
      "grad_norm": 0.705977737903595,
      "learning_rate": 2.8513333854696177e-05,
      "loss": 1.225,
      "step": 11000
    },
    {
      "epoch": 0.43386491557223267,
      "grad_norm": 1.4560986757278442,
      "learning_rate": 2.8317822788769842e-05,
      "loss": 1.2255,
      "step": 11100
    },
    {
      "epoch": 0.4377736085053158,
      "grad_norm": 1.0365822315216064,
      "learning_rate": 2.8122311722843514e-05,
      "loss": 1.2228,
      "step": 11200
    },
    {
      "epoch": 0.441682301438399,
      "grad_norm": 0.9956429600715637,
      "learning_rate": 2.7926800656917183e-05,
      "loss": 1.2185,
      "step": 11300
    },
    {
      "epoch": 0.4455909943714822,
      "grad_norm": 0.9062424302101135,
      "learning_rate": 2.773128959099085e-05,
      "loss": 1.2229,
      "step": 11400
    },
    {
      "epoch": 0.44949968730456535,
      "grad_norm": 1.64096999168396,
      "learning_rate": 2.753577852506452e-05,
      "loss": 1.2186,
      "step": 11500
    },
    {
      "epoch": 0.4534083802376485,
      "grad_norm": 1.88444983959198,
      "learning_rate": 2.7340267459138186e-05,
      "loss": 1.2206,
      "step": 11600
    },
    {
      "epoch": 0.4573170731707317,
      "grad_norm": 1.2047277688980103,
      "learning_rate": 2.7144756393211855e-05,
      "loss": 1.2166,
      "step": 11700
    },
    {
      "epoch": 0.4612257661038149,
      "grad_norm": 1.3528029918670654,
      "learning_rate": 2.6949245327285527e-05,
      "loss": 1.2052,
      "step": 11800
    },
    {
      "epoch": 0.46513445903689804,
      "grad_norm": 0.8786093592643738,
      "learning_rate": 2.6753734261359193e-05,
      "loss": 1.2166,
      "step": 11900
    },
    {
      "epoch": 0.46904315196998125,
      "grad_norm": 1.253136157989502,
      "learning_rate": 2.655822319543286e-05,
      "loss": 1.2184,
      "step": 12000
    },
    {
      "epoch": 0.4729518449030644,
      "grad_norm": 1.228351354598999,
      "learning_rate": 2.6362712129506534e-05,
      "loss": 1.2095,
      "step": 12100
    },
    {
      "epoch": 0.47686053783614757,
      "grad_norm": 1.020992398262024,
      "learning_rate": 2.61672010635802e-05,
      "loss": 1.2028,
      "step": 12200
    },
    {
      "epoch": 0.4807692307692308,
      "grad_norm": 1.2239114046096802,
      "learning_rate": 2.597168999765387e-05,
      "loss": 1.2082,
      "step": 12300
    },
    {
      "epoch": 0.48467792370231394,
      "grad_norm": 0.8391302824020386,
      "learning_rate": 2.5776178931727533e-05,
      "loss": 1.2115,
      "step": 12400
    },
    {
      "epoch": 0.4885866166353971,
      "grad_norm": 1.241330623626709,
      "learning_rate": 2.5580667865801205e-05,
      "loss": 1.2026,
      "step": 12500
    },
    {
      "epoch": 0.4924953095684803,
      "grad_norm": 1.6619151830673218,
      "learning_rate": 2.5385156799874878e-05,
      "loss": 1.2068,
      "step": 12600
    },
    {
      "epoch": 0.49640400250156347,
      "grad_norm": 1.0257009267807007,
      "learning_rate": 2.5189645733948543e-05,
      "loss": 1.2058,
      "step": 12700
    },
    {
      "epoch": 0.5003126954346466,
      "grad_norm": 1.6809911727905273,
      "learning_rate": 2.4994134668022212e-05,
      "loss": 1.2003,
      "step": 12800
    },
    {
      "epoch": 0.5042213883677298,
      "grad_norm": 0.6175216436386108,
      "learning_rate": 2.479862360209588e-05,
      "loss": 1.2084,
      "step": 12900
    },
    {
      "epoch": 0.508130081300813,
      "grad_norm": 1.0972894430160522,
      "learning_rate": 2.460311253616955e-05,
      "loss": 1.1951,
      "step": 13000
    },
    {
      "epoch": 0.5120387742338962,
      "grad_norm": 1.5631442070007324,
      "learning_rate": 2.4407601470243215e-05,
      "loss": 1.2004,
      "step": 13100
    },
    {
      "epoch": 0.5159474671669794,
      "grad_norm": 1.4894869327545166,
      "learning_rate": 2.4212090404316887e-05,
      "loss": 1.1905,
      "step": 13200
    },
    {
      "epoch": 0.5198561601000625,
      "grad_norm": 1.101841926574707,
      "learning_rate": 2.4016579338390556e-05,
      "loss": 1.1967,
      "step": 13300
    },
    {
      "epoch": 0.5237648530331457,
      "grad_norm": 1.0999876260757446,
      "learning_rate": 2.382106827246422e-05,
      "loss": 1.1895,
      "step": 13400
    },
    {
      "epoch": 0.5276735459662288,
      "grad_norm": 1.0898387432098389,
      "learning_rate": 2.362555720653789e-05,
      "loss": 1.1902,
      "step": 13500
    },
    {
      "epoch": 0.5315822388993121,
      "grad_norm": 1.2711913585662842,
      "learning_rate": 2.343004614061156e-05,
      "loss": 1.1957,
      "step": 13600
    },
    {
      "epoch": 0.5354909318323953,
      "grad_norm": 1.1854056119918823,
      "learning_rate": 2.3234535074685228e-05,
      "loss": 1.1905,
      "step": 13700
    },
    {
      "epoch": 0.5393996247654784,
      "grad_norm": 1.2872633934020996,
      "learning_rate": 2.3039024008758896e-05,
      "loss": 1.1908,
      "step": 13800
    },
    {
      "epoch": 0.5433083176985616,
      "grad_norm": 1.082897663116455,
      "learning_rate": 2.2843512942832565e-05,
      "loss": 1.1822,
      "step": 13900
    },
    {
      "epoch": 0.5472170106316447,
      "grad_norm": 1.7656800746917725,
      "learning_rate": 2.2648001876906234e-05,
      "loss": 1.1883,
      "step": 14000
    },
    {
      "epoch": 0.551125703564728,
      "grad_norm": 2.594040632247925,
      "learning_rate": 2.2452490810979903e-05,
      "loss": 1.1881,
      "step": 14100
    },
    {
      "epoch": 0.5550343964978112,
      "grad_norm": 1.6944106817245483,
      "learning_rate": 2.225697974505357e-05,
      "loss": 1.1869,
      "step": 14200
    },
    {
      "epoch": 0.5589430894308943,
      "grad_norm": 1.4870433807373047,
      "learning_rate": 2.206146867912724e-05,
      "loss": 1.1862,
      "step": 14300
    },
    {
      "epoch": 0.5628517823639775,
      "grad_norm": 2.2079107761383057,
      "learning_rate": 2.1865957613200906e-05,
      "loss": 1.1794,
      "step": 14400
    },
    {
      "epoch": 0.5667604752970606,
      "grad_norm": 1.419458270072937,
      "learning_rate": 2.1670446547274575e-05,
      "loss": 1.1798,
      "step": 14500
    },
    {
      "epoch": 0.5706691682301438,
      "grad_norm": 1.2365930080413818,
      "learning_rate": 2.1474935481348247e-05,
      "loss": 1.1827,
      "step": 14600
    },
    {
      "epoch": 0.5745778611632271,
      "grad_norm": 1.2718974351882935,
      "learning_rate": 2.1279424415421916e-05,
      "loss": 1.1805,
      "step": 14700
    },
    {
      "epoch": 0.5784865540963102,
      "grad_norm": 1.7296931743621826,
      "learning_rate": 2.108391334949558e-05,
      "loss": 1.1797,
      "step": 14800
    },
    {
      "epoch": 0.5823952470293934,
      "grad_norm": 0.8652032613754272,
      "learning_rate": 2.088840228356925e-05,
      "loss": 1.1719,
      "step": 14900
    },
    {
      "epoch": 0.5863039399624765,
      "grad_norm": 2.346026659011841,
      "learning_rate": 2.0692891217642922e-05,
      "loss": 1.1847,
      "step": 15000
    },
    {
      "epoch": 0.5902126328955597,
      "grad_norm": 1.951084852218628,
      "learning_rate": 2.0497380151716587e-05,
      "loss": 1.1794,
      "step": 15100
    },
    {
      "epoch": 0.5941213258286429,
      "grad_norm": 1.3734288215637207,
      "learning_rate": 2.0301869085790256e-05,
      "loss": 1.1749,
      "step": 15200
    },
    {
      "epoch": 0.5980300187617261,
      "grad_norm": 2.9074900150299072,
      "learning_rate": 2.0106358019863925e-05,
      "loss": 1.1729,
      "step": 15300
    },
    {
      "epoch": 0.6019387116948093,
      "grad_norm": 2.278609275817871,
      "learning_rate": 1.9910846953937594e-05,
      "loss": 1.1703,
      "step": 15400
    },
    {
      "epoch": 0.6058474046278924,
      "grad_norm": 1.291671633720398,
      "learning_rate": 1.9715335888011263e-05,
      "loss": 1.1621,
      "step": 15500
    },
    {
      "epoch": 0.6097560975609756,
      "grad_norm": 1.9885281324386597,
      "learning_rate": 1.951982482208493e-05,
      "loss": 1.1636,
      "step": 15600
    },
    {
      "epoch": 0.6136647904940588,
      "grad_norm": 1.5937877893447876,
      "learning_rate": 1.93243137561586e-05,
      "loss": 1.1701,
      "step": 15700
    },
    {
      "epoch": 0.6175734834271419,
      "grad_norm": 2.537714719772339,
      "learning_rate": 1.9128802690232266e-05,
      "loss": 1.1659,
      "step": 15800
    },
    {
      "epoch": 0.6214821763602252,
      "grad_norm": 1.8786745071411133,
      "learning_rate": 1.8933291624305938e-05,
      "loss": 1.163,
      "step": 15900
    },
    {
      "epoch": 0.6253908692933083,
      "grad_norm": 1.65304696559906,
      "learning_rate": 1.8737780558379607e-05,
      "loss": 1.1624,
      "step": 16000
    }
  ],
  "logging_steps": 100,
  "max_steps": 25584,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.474282180608e+16,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": null
}