{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2286,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_loss": 2.3424127101898193,
      "eval_runtime": 1.1989,
      "eval_samples_per_second": 16.682,
      "eval_steps_per_second": 4.171,
      "step": 0
    },
    {
      "epoch": 0.0004375410194705754,
      "grad_norm": 11.591988563537598,
      "learning_rate": 3.6231884057971015e-07,
      "loss": 1.7217,
      "step": 1
    },
    {
      "epoch": 0.004375410194705754,
      "grad_norm": 7.750391006469727,
      "learning_rate": 3.6231884057971017e-06,
      "loss": 1.6447,
      "step": 10
    },
    {
      "epoch": 0.008750820389411508,
      "grad_norm": 4.005843162536621,
      "learning_rate": 7.246376811594203e-06,
      "loss": 1.5732,
      "step": 20
    },
    {
      "epoch": 0.01312623058411726,
      "grad_norm": 2.8733792304992676,
      "learning_rate": 1.0869565217391305e-05,
      "loss": 1.4077,
      "step": 30
    },
    {
      "epoch": 0.017501640778823015,
      "grad_norm": 2.6543965339660645,
      "learning_rate": 1.4492753623188407e-05,
      "loss": 1.3353,
      "step": 40
    },
    {
      "epoch": 0.021877050973528767,
      "grad_norm": 2.3828840255737305,
      "learning_rate": 1.8115942028985507e-05,
      "loss": 1.2485,
      "step": 50
    },
    {
      "epoch": 0.02625246116823452,
      "grad_norm": 2.1744871139526367,
      "learning_rate": 2.173913043478261e-05,
      "loss": 1.2685,
      "step": 60
    },
    {
      "epoch": 0.030627871362940276,
      "grad_norm": 2.292287588119507,
      "learning_rate": 2.5362318840579714e-05,
      "loss": 1.1884,
      "step": 70
    },
    {
      "epoch": 0.03500328155764603,
      "grad_norm": 2.3060696125030518,
      "learning_rate": 2.8985507246376814e-05,
      "loss": 1.2874,
      "step": 80
    },
    {
      "epoch": 0.039378691752351785,
      "grad_norm": 2.0335893630981445,
      "learning_rate": 3.260869565217392e-05,
      "loss": 1.2227,
      "step": 90
    },
    {
      "epoch": 0.04375410194705753,
      "grad_norm": 2.4152441024780273,
      "learning_rate": 3.6231884057971014e-05,
      "loss": 1.192,
      "step": 100
    },
    {
      "epoch": 0.04812951214176329,
      "grad_norm": 2.216953992843628,
      "learning_rate": 3.985507246376812e-05,
      "loss": 1.1883,
      "step": 110
    },
    {
      "epoch": 0.05250492233646904,
      "grad_norm": 2.3724114894866943,
      "learning_rate": 4.347826086956522e-05,
      "loss": 1.1811,
      "step": 120
    },
    {
      "epoch": 0.0568803325311748,
      "grad_norm": 2.238722801208496,
      "learning_rate": 4.710144927536232e-05,
      "loss": 1.168,
      "step": 130
    },
    {
      "epoch": 0.06125574272588055,
      "grad_norm": 2.2097883224487305,
      "learning_rate": 4.999997487707664e-05,
      "loss": 1.1848,
      "step": 140
    },
    {
      "epoch": 0.0656311529205863,
      "grad_norm": 2.2063302993774414,
      "learning_rate": 4.999909558006067e-05,
      "loss": 1.2155,
      "step": 150
    },
    {
      "epoch": 0.07000656311529206,
      "grad_norm": 1.8397595882415771,
      "learning_rate": 4.9996960187368994e-05,
      "loss": 1.1507,
      "step": 160
    },
    {
      "epoch": 0.07438197330999781,
      "grad_norm": 1.8552066087722778,
      "learning_rate": 4.9993568806295786e-05,
      "loss": 1.2109,
      "step": 170
    },
    {
      "epoch": 0.07875738350470357,
      "grad_norm": 2.173321485519409,
      "learning_rate": 4.998892160724318e-05,
      "loss": 1.1784,
      "step": 180
    },
    {
      "epoch": 0.08313279369940932,
      "grad_norm": 2.0949854850769043,
      "learning_rate": 4.998301882371268e-05,
      "loss": 1.1967,
      "step": 190
    },
    {
      "epoch": 0.08750820389411507,
      "grad_norm": 1.8619498014450073,
      "learning_rate": 4.997586075229346e-05,
      "loss": 1.1761,
      "step": 200
    },
    {
      "epoch": 0.09188361408882083,
      "grad_norm": 1.9095852375030518,
      "learning_rate": 4.996744775264743e-05,
      "loss": 1.15,
      "step": 210
    },
    {
      "epoch": 0.09625902428352658,
      "grad_norm": 2.0714683532714844,
      "learning_rate": 4.995778024749118e-05,
      "loss": 1.1477,
      "step": 220
    },
    {
      "epoch": 0.10063443447823234,
      "grad_norm": 1.9555009603500366,
      "learning_rate": 4.9946858722574743e-05,
      "loss": 1.1602,
      "step": 230
    },
    {
      "epoch": 0.10500984467293809,
      "grad_norm": 1.9120110273361206,
      "learning_rate": 4.993468372665718e-05,
      "loss": 1.1516,
      "step": 240
    },
    {
      "epoch": 0.10938525486764385,
      "grad_norm": 1.8210362195968628,
      "learning_rate": 4.9921255871479e-05,
      "loss": 1.1404,
      "step": 250
    },
    {
      "epoch": 0.1137606650623496,
      "grad_norm": 1.8263871669769287,
      "learning_rate": 4.990657583173146e-05,
      "loss": 1.1629,
      "step": 260
    },
    {
      "epoch": 0.11813607525705534,
      "grad_norm": 1.906760811805725,
      "learning_rate": 4.989064434502262e-05,
      "loss": 1.1364,
      "step": 270
    },
    {
      "epoch": 0.1225114854517611,
      "grad_norm": 1.7693920135498047,
      "learning_rate": 4.9873462211840296e-05,
      "loss": 1.1396,
      "step": 280
    },
    {
      "epoch": 0.12688689564646685,
      "grad_norm": 1.8662511110305786,
      "learning_rate": 4.985503029551184e-05,
      "loss": 1.1859,
      "step": 290
    },
    {
      "epoch": 0.1312623058411726,
      "grad_norm": 1.7560139894485474,
      "learning_rate": 4.983534952216079e-05,
      "loss": 1.1419,
      "step": 300
    },
    {
      "epoch": 0.13563771603587838,
      "grad_norm": 1.8666057586669922,
      "learning_rate": 4.981442088066029e-05,
      "loss": 1.1686,
      "step": 310
    },
    {
      "epoch": 0.14001312623058412,
      "grad_norm": 1.7406998872756958,
      "learning_rate": 4.9792245422583406e-05,
      "loss": 1.1275,
      "step": 320
    },
    {
      "epoch": 0.14438853642528987,
      "grad_norm": 1.717785120010376,
      "learning_rate": 4.9768824262150344e-05,
      "loss": 1.1236,
      "step": 330
    },
    {
      "epoch": 0.14876394661999562,
      "grad_norm": 1.7829722166061401,
      "learning_rate": 4.9744158576172406e-05,
      "loss": 1.1297,
      "step": 340
    },
    {
      "epoch": 0.15313935681470137,
      "grad_norm": 1.8614799976348877,
      "learning_rate": 4.971824960399289e-05,
      "loss": 1.2018,
      "step": 350
    },
    {
      "epoch": 0.15751476700940714,
      "grad_norm": 1.9301059246063232,
      "learning_rate": 4.9691098647424806e-05,
      "loss": 1.1491,
      "step": 360
    },
    {
      "epoch": 0.1618901772041129,
      "grad_norm": 1.7714897394180298,
      "learning_rate": 4.9662707070685476e-05,
      "loss": 1.1393,
      "step": 370
    },
    {
      "epoch": 0.16626558739881864,
      "grad_norm": 1.7732449769973755,
      "learning_rate": 4.963307630032797e-05,
      "loss": 1.1029,
      "step": 380
    },
    {
      "epoch": 0.17064099759352439,
      "grad_norm": 1.6674808263778687,
      "learning_rate": 4.960220782516946e-05,
      "loss": 1.1385,
      "step": 390
    },
    {
      "epoch": 0.17501640778823013,
      "grad_norm": 1.8136391639709473,
      "learning_rate": 4.957010319621638e-05,
      "loss": 1.0898,
      "step": 400
    },
    {
      "epoch": 0.1793918179829359,
      "grad_norm": 1.7509042024612427,
      "learning_rate": 4.95367640265865e-05,
      "loss": 1.1914,
      "step": 410
    },
    {
      "epoch": 0.18376722817764166,
      "grad_norm": 1.7148741483688354,
      "learning_rate": 4.950219199142791e-05,
      "loss": 1.1228,
      "step": 420
    },
    {
      "epoch": 0.1881426383723474,
      "grad_norm": 1.8227111101150513,
      "learning_rate": 4.9466388827834795e-05,
      "loss": 1.1556,
      "step": 430
    },
    {
      "epoch": 0.19251804856705315,
      "grad_norm": 1.6873571872711182,
      "learning_rate": 4.9429356334760205e-05,
      "loss": 1.1551,
      "step": 440
    },
    {
      "epoch": 0.19689345876175893,
      "grad_norm": 1.6948832273483276,
      "learning_rate": 4.9391096372925626e-05,
      "loss": 1.1085,
      "step": 450
    },
    {
      "epoch": 0.20126886895646467,
      "grad_norm": 1.7749884128570557,
      "learning_rate": 4.9351610864727504e-05,
      "loss": 1.1421,
      "step": 460
    },
    {
      "epoch": 0.20564427915117042,
      "grad_norm": 1.607061743736267,
      "learning_rate": 4.9310901794140665e-05,
      "loss": 1.155,
      "step": 470
    },
    {
      "epoch": 0.21001968934587617,
      "grad_norm": 1.6496872901916504,
      "learning_rate": 4.92689712066186e-05,
      "loss": 1.1552,
      "step": 480
    },
    {
      "epoch": 0.21439509954058192,
      "grad_norm": 1.5627405643463135,
      "learning_rate": 4.922582120899072e-05,
      "loss": 1.1419,
      "step": 490
    },
    {
      "epoch": 0.2187705097352877,
      "grad_norm": 1.7042529582977295,
      "learning_rate": 4.918145396935648e-05,
      "loss": 1.1652,
      "step": 500
    },
    {
      "epoch": 0.22314591992999344,
      "grad_norm": 1.5909007787704468,
      "learning_rate": 4.913587171697643e-05,
      "loss": 1.1294,
      "step": 510
    },
    {
      "epoch": 0.2275213301246992,
      "grad_norm": 1.7357898950576782,
      "learning_rate": 4.908907674216023e-05,
      "loss": 1.1119,
      "step": 520
    },
    {
      "epoch": 0.23189674031940494,
      "grad_norm": 1.6394789218902588,
      "learning_rate": 4.9041071396151585e-05,
      "loss": 1.1626,
      "step": 530
    },
    {
      "epoch": 0.23627215051411068,
      "grad_norm": 1.5801527500152588,
      "learning_rate": 4.899185809101003e-05,
      "loss": 1.1211,
      "step": 540
    },
    {
      "epoch": 0.24064756070881646,
      "grad_norm": 1.8420217037200928,
      "learning_rate": 4.894143929948982e-05,
      "loss": 1.1103,
      "step": 550
    },
    {
      "epoch": 0.2450229709035222,
      "grad_norm": 1.671028971672058,
      "learning_rate": 4.8889817554915645e-05,
      "loss": 1.1461,
      "step": 560
    },
    {
      "epoch": 0.24939838109822796,
      "grad_norm": 1.6634349822998047,
      "learning_rate": 4.8836995451055344e-05,
      "loss": 1.1562,
      "step": 570
    },
    {
      "epoch": 0.2537737912929337,
      "grad_norm": 1.5025454759597778,
      "learning_rate": 4.878297564198959e-05,
      "loss": 1.0946,
      "step": 580
    },
    {
      "epoch": 0.2581492014876395,
      "grad_norm": 1.7138441801071167,
      "learning_rate": 4.872776084197851e-05,
      "loss": 1.1093,
      "step": 590
    },
    {
      "epoch": 0.2625246116823452,
      "grad_norm": 1.668101191520691,
      "learning_rate": 4.867135382532533e-05,
      "loss": 1.153,
      "step": 600
    },
    {
      "epoch": 0.266900021877051,
      "grad_norm": 1.706669569015503,
      "learning_rate": 4.861375742623697e-05,
      "loss": 1.1281,
      "step": 610
    },
    {
      "epoch": 0.27127543207175675,
      "grad_norm": 2.0017876625061035,
      "learning_rate": 4.855497453868164e-05,
      "loss": 1.1056,
      "step": 620
    },
    {
      "epoch": 0.27565084226646247,
      "grad_norm": 1.611232042312622,
      "learning_rate": 4.849500811624343e-05,
      "loss": 1.1271,
      "step": 630
    },
    {
      "epoch": 0.28002625246116825,
      "grad_norm": 1.463722586631775,
      "learning_rate": 4.8433861171973896e-05,
      "loss": 1.1402,
      "step": 640
    },
    {
      "epoch": 0.28440166265587397,
      "grad_norm": 1.5447578430175781,
      "learning_rate": 4.837153677824067e-05,
      "loss": 1.1493,
      "step": 650
    },
    {
      "epoch": 0.28877707285057974,
      "grad_norm": 1.496686339378357,
      "learning_rate": 4.830803806657311e-05,
      "loss": 1.1333,
      "step": 660
    },
    {
      "epoch": 0.2931524830452855,
      "grad_norm": 1.6186118125915527,
      "learning_rate": 4.8243368227504924e-05,
      "loss": 1.1545,
      "step": 670
    },
    {
      "epoch": 0.29752789323999124,
      "grad_norm": 1.5591745376586914,
      "learning_rate": 4.817753051041387e-05,
      "loss": 1.1189,
      "step": 680
    },
    {
      "epoch": 0.301903303434697,
      "grad_norm": 1.8602070808410645,
      "learning_rate": 4.811052822335849e-05,
      "loss": 1.1486,
      "step": 690
    },
    {
      "epoch": 0.30627871362940273,
      "grad_norm": 1.5043679475784302,
      "learning_rate": 4.80423647329119e-05,
      "loss": 1.0961,
      "step": 700
    },
    {
      "epoch": 0.3106541238241085,
      "grad_norm": 1.6095219850540161,
      "learning_rate": 4.797304346399263e-05,
      "loss": 1.141,
      "step": 710
    },
    {
      "epoch": 0.3150295340188143,
      "grad_norm": 1.628132939338684,
      "learning_rate": 4.790256789969252e-05,
      "loss": 1.1784,
      "step": 720
    },
    {
      "epoch": 0.31940494421352,
      "grad_norm": 1.5121166706085205,
      "learning_rate": 4.783094158110174e-05,
      "loss": 1.1193,
      "step": 730
    },
    {
      "epoch": 0.3237803544082258,
      "grad_norm": 1.5826560258865356,
      "learning_rate": 4.775816810713086e-05,
      "loss": 1.1206,
      "step": 740
    },
    {
      "epoch": 0.3281557646029315,
      "grad_norm": 1.5174522399902344,
      "learning_rate": 4.768425113433e-05,
      "loss": 1.1566,
      "step": 750
    },
    {
      "epoch": 0.3325311747976373,
      "grad_norm": 1.6296701431274414,
      "learning_rate": 4.7609194376705105e-05,
      "loss": 1.1388,
      "step": 760
    },
    {
      "epoch": 0.33690658499234305,
      "grad_norm": 1.798865556716919,
      "learning_rate": 4.753300160553136e-05,
      "loss": 1.1889,
      "step": 770
    },
    {
      "epoch": 0.34128199518704877,
      "grad_norm": 1.6122382879257202,
      "learning_rate": 4.745567664916368e-05,
      "loss": 1.125,
      "step": 780
    },
    {
      "epoch": 0.34565740538175455,
      "grad_norm": 1.6329957246780396,
      "learning_rate": 4.737722339284436e-05,
      "loss": 1.121,
      "step": 790
    },
    {
      "epoch": 0.35003281557646027,
      "grad_norm": 1.697094440460205,
      "learning_rate": 4.7297645778507835e-05,
      "loss": 1.0644,
      "step": 800
    },
    {
      "epoch": 0.35440822577116604,
      "grad_norm": 1.6578021049499512,
      "learning_rate": 4.721694780458266e-05,
      "loss": 1.1579,
      "step": 810
    },
    {
      "epoch": 0.3587836359658718,
      "grad_norm": 1.478832483291626,
      "learning_rate": 4.7135133525790544e-05,
      "loss": 1.1512,
      "step": 820
    },
    {
      "epoch": 0.36315904616057754,
      "grad_norm": 1.513938546180725,
      "learning_rate": 4.7052207052942685e-05,
      "loss": 1.0791,
      "step": 830
    },
    {
      "epoch": 0.3675344563552833,
      "grad_norm": 1.5836193561553955,
      "learning_rate": 4.6968172552733185e-05,
      "loss": 1.1259,
      "step": 840
    },
    {
      "epoch": 0.3719098665499891,
      "grad_norm": 1.617882490158081,
      "learning_rate": 4.688303424752969e-05,
      "loss": 1.1001,
      "step": 850
    },
    {
      "epoch": 0.3762852767446948,
      "grad_norm": 1.4709640741348267,
      "learning_rate": 4.679679641516122e-05,
      "loss": 1.1085,
      "step": 860
    },
    {
      "epoch": 0.3806606869394006,
      "grad_norm": 1.4871046543121338,
      "learning_rate": 4.670946338870329e-05,
      "loss": 1.0863,
      "step": 870
    },
    {
      "epoch": 0.3850360971341063,
      "grad_norm": 1.5609277486801147,
      "learning_rate": 4.6621039556260095e-05,
      "loss": 1.1255,
      "step": 880
    },
    {
      "epoch": 0.3894115073288121,
      "grad_norm": 1.513523817062378,
      "learning_rate": 4.653152936074413e-05,
      "loss": 1.1101,
      "step": 890
    },
    {
      "epoch": 0.39378691752351785,
      "grad_norm": 1.6045000553131104,
      "learning_rate": 4.644093729965287e-05,
      "loss": 1.1041,
      "step": 900
    },
    {
      "epoch": 0.3981623277182236,
      "grad_norm": 1.5579705238342285,
      "learning_rate": 4.634926792484284e-05,
      "loss": 1.1592,
      "step": 910
    },
    {
      "epoch": 0.40253773791292935,
      "grad_norm": 1.4306694269180298,
      "learning_rate": 4.625652584230089e-05,
      "loss": 1.1465,
      "step": 920
    },
    {
      "epoch": 0.40691314810763507,
      "grad_norm": 1.3425192832946777,
      "learning_rate": 4.616271571191273e-05,
      "loss": 1.1555,
      "step": 930
    },
    {
      "epoch": 0.41128855830234085,
      "grad_norm": 1.5686333179473877,
      "learning_rate": 4.606784224722887e-05,
      "loss": 1.1394,
      "step": 940
    },
    {
      "epoch": 0.4156639684970466,
      "grad_norm": 1.4096781015396118,
      "learning_rate": 4.5971910215227684e-05,
      "loss": 1.0751,
      "step": 950
    },
    {
      "epoch": 0.42003937869175234,
      "grad_norm": 1.3961308002471924,
      "learning_rate": 4.5874924436075996e-05,
      "loss": 1.1281,
      "step": 960
    },
    {
      "epoch": 0.4244147888864581,
      "grad_norm": 1.4717061519622803,
      "learning_rate": 4.577688978288681e-05,
      "loss": 1.1441,
      "step": 970
    },
    {
      "epoch": 0.42879019908116384,
      "grad_norm": 1.4169813394546509,
      "learning_rate": 4.567781118147448e-05,
      "loss": 1.1381,
      "step": 980
    },
    {
      "epoch": 0.4331656092758696,
      "grad_norm": 1.4888379573822021,
      "learning_rate": 4.557769361010721e-05,
      "loss": 1.0826,
      "step": 990
    },
    {
      "epoch": 0.4375410194705754,
      "grad_norm": 1.572547435760498,
      "learning_rate": 4.547654209925693e-05,
      "loss": 1.1131,
      "step": 1000
    },
    {
      "epoch": 0.4419164296652811,
      "grad_norm": 1.5723587274551392,
      "learning_rate": 4.5374361731346526e-05,
      "loss": 1.1088,
      "step": 1010
    },
    {
      "epoch": 0.4462918398599869,
      "grad_norm": 1.5646806955337524,
      "learning_rate": 4.5271157640494456e-05,
      "loss": 1.1313,
      "step": 1020
    },
    {
      "epoch": 0.4506672500546926,
      "grad_norm": 1.6168869733810425,
      "learning_rate": 4.51669350122568e-05,
      "loss": 1.1452,
      "step": 1030
    },
    {
      "epoch": 0.4550426602493984,
      "grad_norm": 1.565577745437622,
      "learning_rate": 4.506169908336673e-05,
      "loss": 1.1473,
      "step": 1040
    },
    {
      "epoch": 0.45941807044410415,
      "grad_norm": 1.6016472578048706,
      "learning_rate": 4.495545514147134e-05,
      "loss": 1.1006,
      "step": 1050
    },
    {
      "epoch": 0.4637934806388099,
      "grad_norm": 1.628869891166687,
      "learning_rate": 4.484820852486598e-05,
      "loss": 1.1327,
      "step": 1060
    },
    {
      "epoch": 0.46816889083351565,
      "grad_norm": 1.4887852668762207,
      "learning_rate": 4.473996462222606e-05,
      "loss": 1.1011,
      "step": 1070
    },
    {
      "epoch": 0.47254430102822137,
      "grad_norm": 1.6507240533828735,
      "learning_rate": 4.463072887233625e-05,
      "loss": 1.15,
      "step": 1080
    },
    {
      "epoch": 0.47691971122292715,
      "grad_norm": 1.5163503885269165,
      "learning_rate": 4.452050676381725e-05,
      "loss": 1.1598,
      "step": 1090
    },
    {
      "epoch": 0.4812951214176329,
      "grad_norm": 1.6033108234405518,
      "learning_rate": 4.440930383484995e-05,
      "loss": 1.118,
      "step": 1100
    },
    {
      "epoch": 0.48567053161233864,
      "grad_norm": 1.4987646341323853,
      "learning_rate": 4.429712567289722e-05,
      "loss": 1.1352,
      "step": 1110
    },
    {
      "epoch": 0.4900459418070444,
      "grad_norm": 1.51768159866333,
      "learning_rate": 4.418397791442314e-05,
      "loss": 1.1433,
      "step": 1120
    },
    {
      "epoch": 0.4944213520017502,
      "grad_norm": 1.3354873657226562,
      "learning_rate": 4.406986624460979e-05,
      "loss": 1.147,
      "step": 1130
    },
    {
      "epoch": 0.4987967621964559,
      "grad_norm": 1.5693355798721313,
      "learning_rate": 4.395479639707158e-05,
      "loss": 1.0808,
      "step": 1140
    },
    {
      "epoch": 0.5031721723911616,
      "grad_norm": 1.5598881244659424,
      "learning_rate": 4.3838774153567195e-05,
      "loss": 1.1247,
      "step": 1150
    },
    {
      "epoch": 0.5075475825858674,
      "grad_norm": 1.4272862672805786,
      "learning_rate": 4.372180534370908e-05,
      "loss": 1.1122,
      "step": 1160
    },
    {
      "epoch": 0.5119229927805732,
      "grad_norm": 1.5112720727920532,
      "learning_rate": 4.3603895844670495e-05,
      "loss": 1.102,
      "step": 1170
    },
    {
      "epoch": 0.516298402975279,
      "grad_norm": 1.589198350906372,
      "learning_rate": 4.3485051580890236e-05,
      "loss": 1.1435,
      "step": 1180
    },
    {
      "epoch": 0.5206738131699847,
      "grad_norm": 1.4863295555114746,
      "learning_rate": 4.336527852377497e-05,
      "loss": 1.1374,
      "step": 1190
    },
    {
      "epoch": 0.5250492233646904,
      "grad_norm": 1.4931763410568237,
      "learning_rate": 4.324458269139916e-05,
      "loss": 1.111,
      "step": 1200
    },
    {
      "epoch": 0.5294246335593962,
      "grad_norm": 1.5186004638671875,
      "learning_rate": 4.3122970148202744e-05,
      "loss": 1.1213,
      "step": 1210
    },
    {
      "epoch": 0.533800043754102,
      "grad_norm": 1.4999425411224365,
      "learning_rate": 4.300044700468635e-05,
      "loss": 1.1388,
      "step": 1220
    },
    {
      "epoch": 0.5381754539488077,
      "grad_norm": 1.5555874109268188,
      "learning_rate": 4.2877019417104334e-05,
      "loss": 1.1028,
      "step": 1230
    },
    {
      "epoch": 0.5425508641435135,
      "grad_norm": 1.4953700304031372,
      "learning_rate": 4.27526935871554e-05,
      "loss": 1.1342,
      "step": 1240
    },
    {
      "epoch": 0.5469262743382192,
      "grad_norm": 1.4581255912780762,
      "learning_rate": 4.262747576167106e-05,
      "loss": 1.0921,
      "step": 1250
    },
    {
      "epoch": 0.5513016845329249,
      "grad_norm": 1.552866816520691,
      "learning_rate": 4.250137223230165e-05,
      "loss": 1.1089,
      "step": 1260
    },
    {
      "epoch": 0.5556770947276307,
      "grad_norm": 1.5412286520004272,
      "learning_rate": 4.2374389335200335e-05,
      "loss": 1.1261,
      "step": 1270
    },
    {
      "epoch": 0.5600525049223365,
      "grad_norm": 1.64509117603302,
      "learning_rate": 4.224653345070466e-05,
      "loss": 1.0668,
      "step": 1280
    },
    {
      "epoch": 0.5644279151170423,
      "grad_norm": 1.4940526485443115,
      "learning_rate": 4.211781100301596e-05,
      "loss": 1.1372,
      "step": 1290
    },
    {
      "epoch": 0.5688033253117479,
      "grad_norm": 1.4879716634750366,
      "learning_rate": 4.198822845987661e-05,
      "loss": 1.1166,
      "step": 1300
    },
    {
      "epoch": 0.5731787355064537,
      "grad_norm": 1.5581066608428955,
      "learning_rate": 4.185779233224506e-05,
      "loss": 1.0663,
      "step": 1310
    },
    {
      "epoch": 0.5775541457011595,
      "grad_norm": 1.5502949953079224,
      "learning_rate": 4.172650917396863e-05,
      "loss": 1.165,
      "step": 1320
    },
    {
      "epoch": 0.5819295558958653,
      "grad_norm": 1.4957443475723267,
      "learning_rate": 4.1594385581454245e-05,
      "loss": 1.1723,
      "step": 1330
    },
    {
      "epoch": 0.586304966090571,
      "grad_norm": 1.4180186986923218,
      "learning_rate": 4.146142819333702e-05,
      "loss": 1.1181,
      "step": 1340
    },
    {
      "epoch": 0.5906803762852767,
      "grad_norm": 1.5630731582641602,
      "learning_rate": 4.132764369014662e-05,
      "loss": 1.1731,
      "step": 1350
    },
    {
      "epoch": 0.5950557864799825,
      "grad_norm": 1.5089911222457886,
      "learning_rate": 4.119303879397169e-05,
      "loss": 1.1199,
      "step": 1360
    },
    {
      "epoch": 0.5994311966746882,
      "grad_norm": 1.4724007844924927,
      "learning_rate": 4.1057620268122004e-05,
      "loss": 1.1108,
      "step": 1370
    },
    {
      "epoch": 0.603806606869394,
      "grad_norm": 1.6644493341445923,
      "learning_rate": 4.092139491678873e-05,
      "loss": 1.0948,
      "step": 1380
    },
    {
      "epoch": 0.6081820170640998,
      "grad_norm": 1.598973035812378,
      "learning_rate": 4.078436958470244e-05,
      "loss": 1.0948,
      "step": 1390
    },
    {
      "epoch": 0.6125574272588055,
      "grad_norm": 1.4203752279281616,
      "learning_rate": 4.064655115678933e-05,
      "loss": 1.0704,
      "step": 1400
    },
    {
      "epoch": 0.6169328374535112,
      "grad_norm": 1.367942214012146,
      "learning_rate": 4.050794655782515e-05,
      "loss": 1.0958,
      "step": 1410
    },
    {
      "epoch": 0.621308247648217,
      "grad_norm": 1.553614616394043,
      "learning_rate": 4.036856275208736e-05,
      "loss": 1.0756,
      "step": 1420
    },
    {
      "epoch": 0.6256836578429228,
      "grad_norm": 1.4172255992889404,
      "learning_rate": 4.022840674300512e-05,
      "loss": 1.1166,
      "step": 1430
    },
    {
      "epoch": 0.6300590680376286,
      "grad_norm": 1.6529884338378906,
      "learning_rate": 4.0087485572807485e-05,
      "loss": 1.069,
      "step": 1440
    },
    {
      "epoch": 0.6344344782323342,
      "grad_norm": 1.5031507015228271,
      "learning_rate": 3.994580632216952e-05,
      "loss": 1.1368,
      "step": 1450
    },
    {
      "epoch": 0.63880988842704,
      "grad_norm": 1.6318256855010986,
      "learning_rate": 3.980337610985651e-05,
      "loss": 1.1184,
      "step": 1460
    },
    {
      "epoch": 0.6431852986217458,
      "grad_norm": 1.5290026664733887,
      "learning_rate": 3.9660202092366316e-05,
      "loss": 1.1252,
      "step": 1470
    },
    {
      "epoch": 0.6475607088164516,
      "grad_norm": 1.4681955575942993,
      "learning_rate": 3.951629146356975e-05,
      "loss": 1.0866,
      "step": 1480
    },
    {
      "epoch": 0.6519361190111573,
      "grad_norm": 1.4883171319961548,
      "learning_rate": 3.937165145434914e-05,
      "loss": 1.1292,
      "step": 1490
    },
    {
      "epoch": 0.656311529205863,
      "grad_norm": 1.5238957405090332,
      "learning_rate": 3.922628933223502e-05,
      "loss": 1.1547,
      "step": 1500
    },
    {
      "epoch": 0.6606869394005688,
      "grad_norm": 1.5324879884719849,
      "learning_rate": 3.9080212401040914e-05,
      "loss": 1.0668,
      "step": 1510
    },
    {
      "epoch": 0.6650623495952745,
      "grad_norm": 1.4942597150802612,
      "learning_rate": 3.8933428000496455e-05,
      "loss": 1.1394,
      "step": 1520
    },
    {
      "epoch": 0.6694377597899803,
      "grad_norm": 1.5268304347991943,
      "learning_rate": 3.878594350587844e-05,
      "loss": 1.0647,
      "step": 1530
    },
    {
      "epoch": 0.6738131699846861,
      "grad_norm": 1.5157787799835205,
      "learning_rate": 3.863776632764043e-05,
      "loss": 1.1521,
      "step": 1540
    },
    {
      "epoch": 0.6781885801793918,
      "grad_norm": 1.5063765048980713,
      "learning_rate": 3.848890391104028e-05,
      "loss": 1.0954,
      "step": 1550
    },
    {
      "epoch": 0.6825639903740975,
      "grad_norm": 1.4812955856323242,
      "learning_rate": 3.833936373576606e-05,
      "loss": 1.1054,
      "step": 1560
    },
    {
      "epoch": 0.6869394005688033,
      "grad_norm": 1.4586228132247925,
      "learning_rate": 3.8189153315560324e-05,
      "loss": 1.1125,
      "step": 1570
    },
    {
      "epoch": 0.6913148107635091,
      "grad_norm": 1.3584743738174438,
      "learning_rate": 3.803828019784248e-05,
      "loss": 1.103,
      "step": 1580
    },
    {
      "epoch": 0.6956902209582149,
      "grad_norm": 1.5196596384048462,
      "learning_rate": 3.788675196332959e-05,
      "loss": 1.1642,
      "step": 1590
    },
    {
      "epoch": 0.7000656311529205,
      "grad_norm": 1.5605676174163818,
      "learning_rate": 3.7734576225655504e-05,
      "loss": 1.1656,
      "step": 1600
    },
    {
      "epoch": 0.7044410413476263,
      "grad_norm": 1.4787431955337524,
      "learning_rate": 3.758176063098829e-05,
      "loss": 1.0855,
      "step": 1610
    },
    {
      "epoch": 0.7088164515423321,
      "grad_norm": 1.4558417797088623,
      "learning_rate": 3.742831285764602e-05,
      "loss": 1.0855,
      "step": 1620
    },
    {
      "epoch": 0.7131918617370379,
      "grad_norm": 1.552485466003418,
      "learning_rate": 3.7274240615710986e-05,
      "loss": 1.0949,
      "step": 1630
    },
    {
      "epoch": 0.7175672719317436,
      "grad_norm": 1.5539509057998657,
      "learning_rate": 3.711955164664235e-05,
      "loss": 1.1285,
      "step": 1640
    },
    {
      "epoch": 0.7219426821264494,
      "grad_norm": 1.4706348180770874,
      "learning_rate": 3.6964253722887075e-05,
      "loss": 1.1041,
      "step": 1650
    },
    {
      "epoch": 0.7263180923211551,
      "grad_norm": 1.460062861442566,
      "learning_rate": 3.680835464748949e-05,
      "loss": 1.0845,
      "step": 1660
    },
    {
      "epoch": 0.7306935025158608,
      "grad_norm": 1.4356744289398193,
      "learning_rate": 3.6651862253699144e-05,
      "loss": 1.0716,
      "step": 1670
    },
    {
      "epoch": 0.7350689127105666,
      "grad_norm": 1.6479544639587402,
      "learning_rate": 3.649478440457726e-05,
      "loss": 1.1017,
      "step": 1680
    },
    {
      "epoch": 0.7394443229052724,
      "grad_norm": 1.4607200622558594,
      "learning_rate": 3.633712899260166e-05,
      "loss": 1.0853,
      "step": 1690
    },
    {
      "epoch": 0.7438197330999782,
      "grad_norm": 1.5237467288970947,
      "learning_rate": 3.617890393927015e-05,
      "loss": 1.1646,
      "step": 1700
    },
    {
      "epoch": 0.7481951432946838,
      "grad_norm": 1.464272379875183,
      "learning_rate": 3.6020117194702555e-05,
      "loss": 1.0863,
      "step": 1710
    },
    {
      "epoch": 0.7525705534893896,
      "grad_norm": 1.440658688545227,
      "learning_rate": 3.586077673724124e-05,
      "loss": 1.128,
      "step": 1720
    },
    {
      "epoch": 0.7569459636840954,
      "grad_norm": 1.5667482614517212,
      "learning_rate": 3.570089057305022e-05,
      "loss": 1.1441,
      "step": 1730
    },
    {
      "epoch": 0.7613213738788012,
      "grad_norm": 1.4171441793441772,
      "learning_rate": 3.5540466735712906e-05,
      "loss": 1.1002,
      "step": 1740
    },
    {
      "epoch": 0.7656967840735069,
      "grad_norm": 1.412792444229126,
      "learning_rate": 3.5379513285828433e-05,
      "loss": 1.0983,
      "step": 1750
    },
    {
      "epoch": 0.7700721942682126,
      "grad_norm": 1.468995213508606,
      "learning_rate": 3.521803831060668e-05,
      "loss": 1.0736,
      "step": 1760
    },
    {
      "epoch": 0.7744476044629184,
      "grad_norm": 1.570688247680664,
      "learning_rate": 3.505604992346187e-05,
      "loss": 1.0935,
      "step": 1770
    },
    {
      "epoch": 0.7788230146576242,
      "grad_norm": 1.4552085399627686,
      "learning_rate": 3.489355626360497e-05,
      "loss": 1.1388,
      "step": 1780
    },
    {
      "epoch": 0.7831984248523299,
      "grad_norm": 1.6429483890533447,
      "learning_rate": 3.473056549563469e-05,
      "loss": 1.1082,
      "step": 1790
    },
    {
      "epoch": 0.7875738350470357,
      "grad_norm": 1.6186859607696533,
      "learning_rate": 3.456708580912725e-05,
      "loss": 1.1182,
      "step": 1800
    },
    {
      "epoch": 0.7919492452417414,
      "grad_norm": 1.3440533876419067,
      "learning_rate": 3.44031254182249e-05,
      "loss": 1.0769,
      "step": 1810
    },
    {
      "epoch": 0.7963246554364471,
      "grad_norm": 1.3511815071105957,
      "learning_rate": 3.42386925612232e-05,
      "loss": 1.0761,
      "step": 1820
    },
    {
      "epoch": 0.8007000656311529,
      "grad_norm": 1.4499377012252808,
      "learning_rate": 3.407379550015707e-05,
      "loss": 1.1344,
      "step": 1830
    },
    {
      "epoch": 0.8050754758258587,
      "grad_norm": 1.4759228229522705,
      "learning_rate": 3.390844252038565e-05,
      "loss": 1.0874,
      "step": 1840
    },
    {
      "epoch": 0.8094508860205645,
      "grad_norm": 1.4827808141708374,
      "learning_rate": 3.3742641930176045e-05,
      "loss": 1.0313,
      "step": 1850
    },
    {
      "epoch": 0.8138262962152701,
      "grad_norm": 1.6069589853286743,
      "learning_rate": 3.3576402060285786e-05,
      "loss": 1.1089,
      "step": 1860
    },
    {
      "epoch": 0.8182017064099759,
      "grad_norm": 1.4358799457550049,
      "learning_rate": 3.3409731263544345e-05,
      "loss": 1.0682,
      "step": 1870
    },
    {
      "epoch": 0.8225771166046817,
      "grad_norm": 1.4865995645523071,
      "learning_rate": 3.324263791443333e-05,
      "loss": 1.0924,
      "step": 1880
    },
    {
      "epoch": 0.8269525267993875,
      "grad_norm": 1.4410614967346191,
      "learning_rate": 3.307513040866584e-05,
      "loss": 1.0754,
      "step": 1890
    },
    {
      "epoch": 0.8313279369940932,
      "grad_norm": 1.3776915073394775,
      "learning_rate": 3.290721716276448e-05,
      "loss": 1.097,
      "step": 1900
    },
    {
      "epoch": 0.8357033471887989,
      "grad_norm": 1.5188299417495728,
      "learning_rate": 3.273890661363855e-05,
      "loss": 1.1616,
      "step": 1910
    },
    {
      "epoch": 0.8400787573835047,
      "grad_norm": 1.5575170516967773,
      "learning_rate": 3.257020721816013e-05,
      "loss": 1.1246,
      "step": 1920
    },
    {
      "epoch": 0.8444541675782105,
      "grad_norm": 1.5193071365356445,
      "learning_rate": 3.240112745273911e-05,
      "loss": 1.1162,
      "step": 1930
    },
    {
      "epoch": 0.8488295777729162,
      "grad_norm": 1.451002836227417,
      "learning_rate": 3.223167581289729e-05,
      "loss": 1.1106,
      "step": 1940
    },
    {
      "epoch": 0.853204987967622,
      "grad_norm": 1.54093599319458,
      "learning_rate": 3.206186081284161e-05,
      "loss": 1.071,
      "step": 1950
    },
    {
      "epoch": 0.8575803981623277,
      "grad_norm": 1.5598371028900146,
      "learning_rate": 3.1891690985036215e-05,
      "loss": 1.0986,
      "step": 1960
    },
    {
      "epoch": 0.8619558083570334,
      "grad_norm": 1.4229456186294556,
      "learning_rate": 3.1721174879773826e-05,
      "loss": 1.0839,
      "step": 1970
    },
    {
      "epoch": 0.8663312185517392,
      "grad_norm": 1.5579663515090942,
      "learning_rate": 3.1550321064746086e-05,
      "loss": 1.0854,
      "step": 1980
    },
    {
      "epoch": 0.870706628746445,
      "grad_norm": 1.4583122730255127,
      "learning_rate": 3.1379138124613105e-05,
      "loss": 1.0916,
      "step": 1990
    },
    {
      "epoch": 0.8750820389411508,
      "grad_norm": 1.577874779701233,
      "learning_rate": 3.120763466057206e-05,
      "loss": 1.061,
      "step": 2000
    },
    {
      "epoch": 0.8794574491358564,
      "grad_norm": 1.5133827924728394,
      "learning_rate": 3.10358192899251e-05,
      "loss": 1.1227,
      "step": 2010
    },
    {
      "epoch": 0.8838328593305622,
      "grad_norm": 1.4519861936569214,
      "learning_rate": 3.086370064564629e-05,
      "loss": 1.086,
      "step": 2020
    },
    {
      "epoch": 0.888208269525268,
      "grad_norm": 1.6563777923583984,
      "learning_rate": 3.069128737594789e-05,
      "loss": 1.1102,
      "step": 2030
    },
    {
      "epoch": 0.8925836797199738,
      "grad_norm": 1.3476353883743286,
      "learning_rate": 3.051858814384581e-05,
      "loss": 1.0804,
      "step": 2040
    },
    {
      "epoch": 0.8969590899146795,
      "grad_norm": 1.501389503479004,
      "learning_rate": 3.0345611626724285e-05,
      "loss": 1.1101,
      "step": 2050
    },
    {
      "epoch": 0.9013345001093852,
      "grad_norm": 1.483244776725769,
      "learning_rate": 3.0172366515899982e-05,
      "loss": 1.0964,
      "step": 2060
    },
    {
      "epoch": 0.905709910304091,
      "grad_norm": 1.4610023498535156,
      "learning_rate": 2.9998861516185173e-05,
      "loss": 1.0953,
      "step": 2070
    },
    {
      "epoch": 0.9100853204987968,
      "grad_norm": 1.4304044246673584,
      "learning_rate": 2.982510534545043e-05,
      "loss": 1.0687,
      "step": 2080
    },
    {
      "epoch": 0.9144607306935025,
      "grad_norm": 1.461287498474121,
      "learning_rate": 2.9651106734186618e-05,
      "loss": 1.0875,
      "step": 2090
    },
    {
      "epoch": 0.9188361408882083,
      "grad_norm": 1.484709620475769,
      "learning_rate": 2.9476874425066105e-05,
      "loss": 1.1122,
      "step": 2100
    },
    {
      "epoch": 0.923211551082914,
      "grad_norm": 1.4673666954040527,
      "learning_rate": 2.9302417172503622e-05,
      "loss": 1.1506,
      "step": 2110
    },
    {
      "epoch": 0.9275869612776197,
      "grad_norm": 1.456113576889038,
      "learning_rate": 2.9127743742216303e-05,
      "loss": 1.0851,
      "step": 2120
    },
    {
      "epoch": 0.9319623714723255,
      "grad_norm": 1.524404525756836,
      "learning_rate": 2.8952862910783267e-05,
      "loss": 1.1047,
      "step": 2130
    },
    {
      "epoch": 0.9363377816670313,
      "grad_norm": 1.3824920654296875,
      "learning_rate": 2.877778346520466e-05,
      "loss": 1.069,
      "step": 2140
    },
    {
      "epoch": 0.9407131918617371,
      "grad_norm": 1.3911488056182861,
      "learning_rate": 2.860251420246009e-05,
      "loss": 1.1128,
      "step": 2150
    },
    {
      "epoch": 0.9450886020564427,
      "grad_norm": 1.6120619773864746,
      "learning_rate": 2.842706392906669e-05,
      "loss": 1.1468,
      "step": 2160
    },
    {
      "epoch": 0.9494640122511485,
      "grad_norm": 1.4727811813354492,
      "learning_rate": 2.825144146063654e-05,
      "loss": 1.1405,
      "step": 2170
    },
    {
      "epoch": 0.9538394224458543,
      "grad_norm": 1.4946926832199097,
      "learning_rate": 2.8075655621433826e-05,
      "loss": 1.0887,
      "step": 2180
    },
    {
      "epoch": 0.9582148326405601,
      "grad_norm": 1.575395107269287,
      "learning_rate": 2.7899715243931375e-05,
      "loss": 1.077,
      "step": 2190
    },
    {
      "epoch": 0.9625902428352658,
      "grad_norm": 1.483463168144226,
      "learning_rate": 2.7723629168366883e-05,
      "loss": 1.1137,
      "step": 2200
    },
    {
      "epoch": 0.9669656530299715,
      "grad_norm": 1.5409109592437744,
      "learning_rate": 2.754740624229875e-05,
      "loss": 1.0749,
      "step": 2210
    },
    {
      "epoch": 0.9713410632246773,
      "grad_norm": 1.4951428174972534,
      "learning_rate": 2.737105532016152e-05,
      "loss": 1.0779,
      "step": 2220
    },
    {
      "epoch": 0.9757164734193831,
      "grad_norm": 1.6109696626663208,
      "learning_rate": 2.719458526282099e-05,
      "loss": 1.1006,
      "step": 2230
    },
    {
      "epoch": 0.9800918836140888,
      "grad_norm": 1.4150285720825195,
      "learning_rate": 2.7018004937128937e-05,
      "loss": 1.0498,
      "step": 2240
    },
    {
      "epoch": 0.9844672938087946,
      "grad_norm": 1.5306379795074463,
      "learning_rate": 2.6841323215477714e-05,
      "loss": 1.0943,
      "step": 2250
    },
    {
      "epoch": 0.9888427040035004,
      "grad_norm": 1.3424153327941895,
      "learning_rate": 2.6664548975354332e-05,
      "loss": 1.079,
      "step": 2260
    },
    {
      "epoch": 0.993218114198206,
      "grad_norm": 1.4553660154342651,
      "learning_rate": 2.6487691098894447e-05,
      "loss": 1.0663,
      "step": 2270
    },
    {
      "epoch": 0.9975935243929118,
      "grad_norm": 1.5150882005691528,
      "learning_rate": 2.6310758472436103e-05,
      "loss": 1.0449,
      "step": 2280
    }
  ],
  "logging_steps": 10,
  "max_steps": 4570,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500.0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.493158922149888e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}