{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9995767195767196,
  "eval_steps": 500,
  "global_step": 2362,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008465608465608466,
      "grad_norm": 1.2536740494709757,
      "learning_rate": 9.999557744171606e-05,
      "loss": 1.6294,
      "step": 10
    },
    {
      "epoch": 0.016931216931216932,
      "grad_norm": 1.0805568114986361,
      "learning_rate": 9.998231054922511e-05,
      "loss": 0.9236,
      "step": 20
    },
    {
      "epoch": 0.025396825396825397,
      "grad_norm": 0.8782195431244562,
      "learning_rate": 9.996020166947136e-05,
      "loss": 0.7603,
      "step": 30
    },
    {
      "epoch": 0.033862433862433865,
      "grad_norm": 0.9267321940966723,
      "learning_rate": 9.99292547135672e-05,
      "loss": 0.6731,
      "step": 40
    },
    {
      "epoch": 0.042328042328042326,
      "grad_norm": 0.8701878795418962,
      "learning_rate": 9.988947515610124e-05,
      "loss": 0.6137,
      "step": 50
    },
    {
      "epoch": 0.050793650793650794,
      "grad_norm": 1.2272042644235537,
      "learning_rate": 9.984087003416997e-05,
      "loss": 0.6346,
      "step": 60
    },
    {
      "epoch": 0.05925925925925926,
      "grad_norm": 1.0686141559390336,
      "learning_rate": 9.978344794613277e-05,
      "loss": 0.6107,
      "step": 70
    },
    {
      "epoch": 0.06772486772486773,
      "grad_norm": 0.9849077305305056,
      "learning_rate": 9.971721905009086e-05,
      "loss": 0.5656,
      "step": 80
    },
    {
      "epoch": 0.0761904761904762,
      "grad_norm": 1.2368756418600437,
      "learning_rate": 9.964219506209039e-05,
      "loss": 0.5726,
      "step": 90
    },
    {
      "epoch": 0.08465608465608465,
      "grad_norm": 1.0374346188421604,
      "learning_rate": 9.955838925404971e-05,
      "loss": 0.5894,
      "step": 100
    },
    {
      "epoch": 0.09312169312169312,
      "grad_norm": 1.258972872704355,
      "learning_rate": 9.946581645141167e-05,
      "loss": 0.534,
      "step": 110
    },
    {
      "epoch": 0.10158730158730159,
      "grad_norm": 1.0466165338353233,
      "learning_rate": 9.936449303052086e-05,
      "loss": 0.5235,
      "step": 120
    },
    {
      "epoch": 0.11005291005291006,
      "grad_norm": 0.9956790587338181,
      "learning_rate": 9.925443691572668e-05,
      "loss": 0.5208,
      "step": 130
    },
    {
      "epoch": 0.11851851851851852,
      "grad_norm": 1.2599033904063475,
      "learning_rate": 9.91356675762124e-05,
      "loss": 0.5355,
      "step": 140
    },
    {
      "epoch": 0.12698412698412698,
      "grad_norm": 1.0353580902393011,
      "learning_rate": 9.900820602255108e-05,
      "loss": 0.5361,
      "step": 150
    },
    {
      "epoch": 0.13544973544973546,
      "grad_norm": 0.9863947603039183,
      "learning_rate": 9.88720748029887e-05,
      "loss": 0.5105,
      "step": 160
    },
    {
      "epoch": 0.1439153439153439,
      "grad_norm": 0.9272966295135331,
      "learning_rate": 9.87272979994554e-05,
      "loss": 0.5286,
      "step": 170
    },
    {
      "epoch": 0.1523809523809524,
      "grad_norm": 0.9357784657446446,
      "learning_rate": 9.857390122330522e-05,
      "loss": 0.4913,
      "step": 180
    },
    {
      "epoch": 0.16084656084656085,
      "grad_norm": 0.9706862272089483,
      "learning_rate": 9.84119116107855e-05,
      "loss": 0.5358,
      "step": 190
    },
    {
      "epoch": 0.1693121693121693,
      "grad_norm": 1.037354134026295,
      "learning_rate": 9.82413578182364e-05,
      "loss": 0.5511,
      "step": 200
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 1.2141517165745577,
      "learning_rate": 9.806227001702135e-05,
      "loss": 0.5099,
      "step": 210
    },
    {
      "epoch": 0.18624338624338624,
      "grad_norm": 1.0430093698701248,
      "learning_rate": 9.787467988818997e-05,
      "loss": 0.5019,
      "step": 220
    },
    {
      "epoch": 0.19470899470899472,
      "grad_norm": 1.0478269154568778,
      "learning_rate": 9.767862061687337e-05,
      "loss": 0.5014,
      "step": 230
    },
    {
      "epoch": 0.20317460317460317,
      "grad_norm": 1.0602153985712557,
      "learning_rate": 9.747412688641372e-05,
      "loss": 0.4933,
      "step": 240
    },
    {
      "epoch": 0.21164021164021163,
      "grad_norm": 1.0783975799525436,
      "learning_rate": 9.726123487222871e-05,
      "loss": 0.4762,
      "step": 250
    },
    {
      "epoch": 0.2201058201058201,
      "grad_norm": 1.008218163473069,
      "learning_rate": 9.703998223541196e-05,
      "loss": 0.4843,
      "step": 260
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 0.9852689728116106,
      "learning_rate": 9.681040811607075e-05,
      "loss": 0.4603,
      "step": 270
    },
    {
      "epoch": 0.23703703703703705,
      "grad_norm": 1.0930039587507208,
      "learning_rate": 9.6572553126402e-05,
      "loss": 0.4859,
      "step": 280
    },
    {
      "epoch": 0.2455026455026455,
      "grad_norm": 0.998867899511094,
      "learning_rate": 9.632645934350792e-05,
      "loss": 0.4764,
      "step": 290
    },
    {
      "epoch": 0.25396825396825395,
      "grad_norm": 1.1270572507755159,
      "learning_rate": 9.607217030195242e-05,
      "loss": 0.5125,
      "step": 300
    },
    {
      "epoch": 0.2624338624338624,
      "grad_norm": 1.0417587165454332,
      "learning_rate": 9.58097309860598e-05,
      "loss": 0.4753,
      "step": 310
    },
    {
      "epoch": 0.2708994708994709,
      "grad_norm": 1.0092231428385308,
      "learning_rate": 9.553918782195688e-05,
      "loss": 0.4654,
      "step": 320
    },
    {
      "epoch": 0.27936507936507937,
      "grad_norm": 1.019655360850703,
      "learning_rate": 9.526058866936013e-05,
      "loss": 0.4897,
      "step": 330
    },
    {
      "epoch": 0.2878306878306878,
      "grad_norm": 1.2136875720692115,
      "learning_rate": 9.497398281310914e-05,
      "loss": 0.4698,
      "step": 340
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 1.0156909584259968,
      "learning_rate": 9.467942095444809e-05,
      "loss": 0.4635,
      "step": 350
    },
    {
      "epoch": 0.3047619047619048,
      "grad_norm": 1.0330410864779689,
      "learning_rate": 9.43769552020565e-05,
      "loss": 0.4955,
      "step": 360
    },
    {
      "epoch": 0.31322751322751324,
      "grad_norm": 1.1555703560002792,
      "learning_rate": 9.406663906283109e-05,
      "loss": 0.4779,
      "step": 370
    },
    {
      "epoch": 0.3216931216931217,
      "grad_norm": 1.0300804678514643,
      "learning_rate": 9.374852743242037e-05,
      "loss": 0.4581,
      "step": 380
    },
    {
      "epoch": 0.33015873015873015,
      "grad_norm": 1.0577163168213004,
      "learning_rate": 9.34226765855134e-05,
      "loss": 0.4578,
      "step": 390
    },
    {
      "epoch": 0.3386243386243386,
      "grad_norm": 1.1528354628813937,
      "learning_rate": 9.308914416588468e-05,
      "loss": 0.4688,
      "step": 400
    },
    {
      "epoch": 0.3470899470899471,
      "grad_norm": 1.041087172048898,
      "learning_rate": 9.27479891761968e-05,
      "loss": 0.4782,
      "step": 410
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 0.9791427857125372,
      "learning_rate": 9.239927196756279e-05,
      "loss": 0.4488,
      "step": 420
    },
    {
      "epoch": 0.364021164021164,
      "grad_norm": 0.9683401420738389,
      "learning_rate": 9.204305422886987e-05,
      "loss": 0.4631,
      "step": 430
    },
    {
      "epoch": 0.3724867724867725,
      "grad_norm": 1.0489264170989323,
      "learning_rate": 9.167939897586647e-05,
      "loss": 0.4493,
      "step": 440
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 1.0133114072958438,
      "learning_rate": 9.130837054001463e-05,
      "loss": 0.461,
      "step": 450
    },
    {
      "epoch": 0.38941798941798944,
      "grad_norm": 1.0851755168134365,
      "learning_rate": 9.09300345571097e-05,
      "loss": 0.469,
      "step": 460
    },
    {
      "epoch": 0.3978835978835979,
      "grad_norm": 1.1009516736250549,
      "learning_rate": 9.054445795566907e-05,
      "loss": 0.4504,
      "step": 470
    },
    {
      "epoch": 0.40634920634920635,
      "grad_norm": 1.0034292869828323,
      "learning_rate": 9.015170894509244e-05,
      "loss": 0.4498,
      "step": 480
    },
    {
      "epoch": 0.4148148148148148,
      "grad_norm": 1.0546642511020514,
      "learning_rate": 8.975185700359542e-05,
      "loss": 0.4334,
      "step": 490
    },
    {
      "epoch": 0.42328042328042326,
      "grad_norm": 0.9211704485733719,
      "learning_rate": 8.93449728659187e-05,
      "loss": 0.4535,
      "step": 500
    },
    {
      "epoch": 0.43174603174603177,
      "grad_norm": 0.9075636796189445,
      "learning_rate": 8.893112851081478e-05,
      "loss": 0.4149,
      "step": 510
    },
    {
      "epoch": 0.4402116402116402,
      "grad_norm": 0.9394762728420766,
      "learning_rate": 8.851039714831492e-05,
      "loss": 0.452,
      "step": 520
    },
    {
      "epoch": 0.4486772486772487,
      "grad_norm": 1.0202540982390778,
      "learning_rate": 8.8082853206778e-05,
      "loss": 0.4491,
      "step": 530
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 1.0209784880401114,
      "learning_rate": 8.764857231972408e-05,
      "loss": 0.4374,
      "step": 540
    },
    {
      "epoch": 0.4656084656084656,
      "grad_norm": 1.004488049664675,
      "learning_rate": 8.72076313124545e-05,
      "loss": 0.459,
      "step": 550
    },
    {
      "epoch": 0.4740740740740741,
      "grad_norm": 1.0341305920498378,
      "learning_rate": 8.676010818846145e-05,
      "loss": 0.4271,
      "step": 560
    },
    {
      "epoch": 0.48253968253968255,
      "grad_norm": 1.0664004827817963,
      "learning_rate": 8.630608211562891e-05,
      "loss": 0.4622,
      "step": 570
    },
    {
      "epoch": 0.491005291005291,
      "grad_norm": 0.9934436898635304,
      "learning_rate": 8.584563341222765e-05,
      "loss": 0.4432,
      "step": 580
    },
    {
      "epoch": 0.49947089947089945,
      "grad_norm": 0.9857548315821839,
      "learning_rate": 8.537884353270677e-05,
      "loss": 0.4639,
      "step": 590
    },
    {
      "epoch": 0.5079365079365079,
      "grad_norm": 1.0178863758450196,
      "learning_rate": 8.490579505328424e-05,
      "loss": 0.4138,
      "step": 600
    },
    {
      "epoch": 0.5164021164021164,
      "grad_norm": 1.038252359047521,
      "learning_rate": 8.442657165733886e-05,
      "loss": 0.419,
      "step": 610
    },
    {
      "epoch": 0.5248677248677248,
      "grad_norm": 0.8829314461573528,
      "learning_rate": 8.394125812060666e-05,
      "loss": 0.4057,
      "step": 620
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 1.0470304353319828,
      "learning_rate": 8.344994029618374e-05,
      "loss": 0.4478,
      "step": 630
    },
    {
      "epoch": 0.5417989417989418,
      "grad_norm": 1.0281770671480395,
      "learning_rate": 8.295270509933862e-05,
      "loss": 0.4382,
      "step": 640
    },
    {
      "epoch": 0.5502645502645502,
      "grad_norm": 0.8764644053490013,
      "learning_rate": 8.24496404921369e-05,
      "loss": 0.4377,
      "step": 650
    },
    {
      "epoch": 0.5587301587301587,
      "grad_norm": 0.9473160233886304,
      "learning_rate": 8.19408354678804e-05,
      "loss": 0.4468,
      "step": 660
    },
    {
      "epoch": 0.5671957671957671,
      "grad_norm": 1.0853088610186052,
      "learning_rate": 8.142638003536413e-05,
      "loss": 0.43,
      "step": 670
    },
    {
      "epoch": 0.5756613756613757,
      "grad_norm": 0.9386265978912677,
      "learning_rate": 8.090636520295348e-05,
      "loss": 0.4475,
      "step": 680
    },
    {
      "epoch": 0.5841269841269842,
      "grad_norm": 1.0740222193078277,
      "learning_rate": 8.038088296248462e-05,
      "loss": 0.4424,
      "step": 690
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 1.0857350308205733,
      "learning_rate": 7.9850026272991e-05,
      "loss": 0.4036,
      "step": 700
    },
    {
      "epoch": 0.6010582010582011,
      "grad_norm": 0.9193145888168105,
      "learning_rate": 7.93138890442586e-05,
      "loss": 0.4511,
      "step": 710
    },
    {
      "epoch": 0.6095238095238096,
      "grad_norm": 0.9551109099001511,
      "learning_rate": 7.877256612021312e-05,
      "loss": 0.3774,
      "step": 720
    },
    {
      "epoch": 0.617989417989418,
      "grad_norm": 0.9570542406982319,
      "learning_rate": 7.822615326214183e-05,
      "loss": 0.4371,
      "step": 730
    },
    {
      "epoch": 0.6264550264550265,
      "grad_norm": 1.0153420855979158,
      "learning_rate": 7.767474713175321e-05,
      "loss": 0.4628,
      "step": 740
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 1.0695763456638891,
      "learning_rate": 7.711844527407724e-05,
      "loss": 0.4238,
      "step": 750
    },
    {
      "epoch": 0.6433862433862434,
      "grad_norm": 1.0044619876891803,
      "learning_rate": 7.655734610020952e-05,
      "loss": 0.4296,
      "step": 760
    },
    {
      "epoch": 0.6518518518518519,
      "grad_norm": 0.9458827543873214,
      "learning_rate": 7.599154886990199e-05,
      "loss": 0.396,
      "step": 770
    },
    {
      "epoch": 0.6603174603174603,
      "grad_norm": 1.074091533938013,
      "learning_rate": 7.542115367400376e-05,
      "loss": 0.4035,
      "step": 780
    },
    {
      "epoch": 0.6687830687830688,
      "grad_norm": 1.0152819418160979,
      "learning_rate": 7.484626141675481e-05,
      "loss": 0.4135,
      "step": 790
    },
    {
      "epoch": 0.6772486772486772,
      "grad_norm": 0.8781718355357042,
      "learning_rate": 7.426697379793572e-05,
      "loss": 0.401,
      "step": 800
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 0.9320133553311277,
      "learning_rate": 7.368339329487677e-05,
      "loss": 0.4257,
      "step": 810
    },
    {
      "epoch": 0.6941798941798942,
      "grad_norm": 1.0257232524871047,
      "learning_rate": 7.309562314432952e-05,
      "loss": 0.4096,
      "step": 820
    },
    {
      "epoch": 0.7026455026455026,
      "grad_norm": 1.000399097214694,
      "learning_rate": 7.250376732420387e-05,
      "loss": 0.4217,
      "step": 830
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 0.8946747414376721,
      "learning_rate": 7.190793053517424e-05,
      "loss": 0.4391,
      "step": 840
    },
    {
      "epoch": 0.7195767195767195,
      "grad_norm": 1.059007250343128,
      "learning_rate": 7.13082181821577e-05,
      "loss": 0.4477,
      "step": 850
    },
    {
      "epoch": 0.728042328042328,
      "grad_norm": 0.9750233036191156,
      "learning_rate": 7.070473635566768e-05,
      "loss": 0.4111,
      "step": 860
    },
    {
      "epoch": 0.7365079365079366,
      "grad_norm": 1.2069014928768744,
      "learning_rate": 7.009759181304619e-05,
      "loss": 0.385,
      "step": 870
    },
    {
      "epoch": 0.744973544973545,
      "grad_norm": 0.9675202014376196,
      "learning_rate": 6.948689195957829e-05,
      "loss": 0.4303,
      "step": 880
    },
    {
      "epoch": 0.7534391534391535,
      "grad_norm": 1.0273630814532357,
      "learning_rate": 6.887274482949185e-05,
      "loss": 0.4197,
      "step": 890
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 0.8936119385994451,
      "learning_rate": 6.825525906684593e-05,
      "loss": 0.428,
      "step": 900
    },
    {
      "epoch": 0.7703703703703704,
      "grad_norm": 0.9839762657140783,
      "learning_rate": 6.763454390631156e-05,
      "loss": 0.4318,
      "step": 910
    },
    {
      "epoch": 0.7788359788359789,
      "grad_norm": 0.9938930890241121,
      "learning_rate": 6.701070915384776e-05,
      "loss": 0.3992,
      "step": 920
    },
    {
      "epoch": 0.7873015873015873,
      "grad_norm": 1.0647555531218702,
      "learning_rate": 6.638386516727656e-05,
      "loss": 0.4055,
      "step": 930
    },
    {
      "epoch": 0.7957671957671958,
      "grad_norm": 0.9510971038124535,
      "learning_rate": 6.575412283676063e-05,
      "loss": 0.3999,
      "step": 940
    },
    {
      "epoch": 0.8042328042328042,
      "grad_norm": 0.9903300056996712,
      "learning_rate": 6.512159356518638e-05,
      "loss": 0.3947,
      "step": 950
    },
    {
      "epoch": 0.8126984126984127,
      "grad_norm": 0.9769472788235855,
      "learning_rate": 6.448638924845662e-05,
      "loss": 0.4052,
      "step": 960
    },
    {
      "epoch": 0.8211640211640212,
      "grad_norm": 1.0442770249636188,
      "learning_rate": 6.384862225569584e-05,
      "loss": 0.3906,
      "step": 970
    },
    {
      "epoch": 0.8296296296296296,
      "grad_norm": 1.1520118825182077,
      "learning_rate": 6.320840540937196e-05,
      "loss": 0.3942,
      "step": 980
    },
    {
      "epoch": 0.8380952380952381,
      "grad_norm": 0.973580565603432,
      "learning_rate": 6.256585196533763e-05,
      "loss": 0.3976,
      "step": 990
    },
    {
      "epoch": 0.8465608465608465,
      "grad_norm": 1.0010129765253053,
      "learning_rate": 6.192107559279513e-05,
      "loss": 0.4132,
      "step": 1000
    },
    {
      "epoch": 0.855026455026455,
      "grad_norm": 1.0317834662297176,
      "learning_rate": 6.127419035418798e-05,
      "loss": 0.4006,
      "step": 1010
    },
    {
      "epoch": 0.8634920634920635,
      "grad_norm": 1.033103546662518,
      "learning_rate": 6.0625310685023006e-05,
      "loss": 0.4015,
      "step": 1020
    },
    {
      "epoch": 0.8719576719576719,
      "grad_norm": 0.9814116522619839,
      "learning_rate": 5.9974551373626456e-05,
      "loss": 0.4042,
      "step": 1030
    },
    {
      "epoch": 0.8804232804232804,
      "grad_norm": 0.9434073558115307,
      "learning_rate": 5.932202754083765e-05,
      "loss": 0.4155,
      "step": 1040
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.9305038165748942,
      "learning_rate": 5.8667854619643926e-05,
      "loss": 0.409,
      "step": 1050
    },
    {
      "epoch": 0.8973544973544973,
      "grad_norm": 0.9961689454529794,
      "learning_rate": 5.8012148334760077e-05,
      "loss": 0.4094,
      "step": 1060
    },
    {
      "epoch": 0.9058201058201059,
      "grad_norm": 1.1061454998502234,
      "learning_rate": 5.735502468215663e-05,
      "loss": 0.3936,
      "step": 1070
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 0.9207561081731734,
      "learning_rate": 5.669659990853975e-05,
      "loss": 0.4096,
      "step": 1080
    },
    {
      "epoch": 0.9227513227513228,
      "grad_norm": 1.105503110809865,
      "learning_rate": 5.603699049078685e-05,
      "loss": 0.3951,
      "step": 1090
    },
    {
      "epoch": 0.9312169312169312,
      "grad_norm": 1.0603861460237525,
      "learning_rate": 5.537631311534176e-05,
      "loss": 0.3881,
      "step": 1100
    },
    {
      "epoch": 0.9396825396825397,
      "grad_norm": 0.9136553453309776,
      "learning_rate": 5.4714684657572456e-05,
      "loss": 0.4021,
      "step": 1110
    },
    {
      "epoch": 0.9481481481481482,
      "grad_norm": 1.0267048307590627,
      "learning_rate": 5.40522221610956e-05,
      "loss": 0.4026,
      "step": 1120
    },
    {
      "epoch": 0.9566137566137566,
      "grad_norm": 1.1963746364272148,
      "learning_rate": 5.338904281707128e-05,
      "loss": 0.3758,
      "step": 1130
    },
    {
      "epoch": 0.9650793650793651,
      "grad_norm": 0.9462817093916603,
      "learning_rate": 5.272526394347156e-05,
      "loss": 0.4041,
      "step": 1140
    },
    {
      "epoch": 0.9735449735449735,
      "grad_norm": 0.9508846912892318,
      "learning_rate": 5.2061002964326655e-05,
      "loss": 0.4102,
      "step": 1150
    },
    {
      "epoch": 0.982010582010582,
      "grad_norm": 0.9999646043761261,
      "learning_rate": 5.139637738895243e-05,
      "loss": 0.3942,
      "step": 1160
    },
    {
      "epoch": 0.9904761904761905,
      "grad_norm": 0.9017617882068949,
      "learning_rate": 5.0731504791162645e-05,
      "loss": 0.3873,
      "step": 1170
    },
    {
      "epoch": 0.9989417989417989,
      "grad_norm": 1.0252778174062445,
      "learning_rate": 5.006650278846991e-05,
      "loss": 0.3813,
      "step": 1180
    },
    {
      "epoch": 1.0074074074074073,
      "grad_norm": 1.0599548491574322,
      "learning_rate": 4.9401489021278844e-05,
      "loss": 0.3662,
      "step": 1190
    },
    {
      "epoch": 1.0158730158730158,
      "grad_norm": 1.218509380892567,
      "learning_rate": 4.873658113207526e-05,
      "loss": 0.3422,
      "step": 1200
    },
    {
      "epoch": 1.0243386243386243,
      "grad_norm": 1.0795258633356928,
      "learning_rate": 4.807189674461489e-05,
      "loss": 0.3386,
      "step": 1210
    },
    {
      "epoch": 1.0328042328042328,
      "grad_norm": 0.9886890483794867,
      "learning_rate": 4.740755344311549e-05,
      "loss": 0.3307,
      "step": 1220
    },
    {
      "epoch": 1.0412698412698413,
      "grad_norm": 1.050177935297077,
      "learning_rate": 4.6743668751455935e-05,
      "loss": 0.3137,
      "step": 1230
    },
    {
      "epoch": 1.0497354497354496,
      "grad_norm": 1.0381034500976596,
      "learning_rate": 4.6080360112385914e-05,
      "loss": 0.3381,
      "step": 1240
    },
    {
      "epoch": 1.0582010582010581,
      "grad_norm": 1.2730980092196418,
      "learning_rate": 4.5417744866750096e-05,
      "loss": 0.3166,
      "step": 1250
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 1.2352467183360096,
      "learning_rate": 4.475594023273024e-05,
      "loss": 0.3264,
      "step": 1260
    },
    {
      "epoch": 1.0751322751322752,
      "grad_norm": 1.0837500411632388,
      "learning_rate": 4.4095063285108975e-05,
      "loss": 0.3207,
      "step": 1270
    },
    {
      "epoch": 1.0835978835978837,
      "grad_norm": 1.0653199477181592,
      "learning_rate": 4.343523093455909e-05,
      "loss": 0.3232,
      "step": 1280
    },
    {
      "epoch": 1.0920634920634922,
      "grad_norm": 1.2017002126624405,
      "learning_rate": 4.277655990696171e-05,
      "loss": 0.3189,
      "step": 1290
    },
    {
      "epoch": 1.1005291005291005,
      "grad_norm": 1.2016108585957874,
      "learning_rate": 4.211916672275722e-05,
      "loss": 0.3305,
      "step": 1300
    },
    {
      "epoch": 1.108994708994709,
      "grad_norm": 0.9929918539213213,
      "learning_rate": 4.14631676763325e-05,
      "loss": 0.3061,
      "step": 1310
    },
    {
      "epoch": 1.1174603174603175,
      "grad_norm": 1.1653749824747823,
      "learning_rate": 4.080867881544826e-05,
      "loss": 0.3182,
      "step": 1320
    },
    {
      "epoch": 1.125925925925926,
      "grad_norm": 1.2857542752538433,
      "learning_rate": 4.0155815920709825e-05,
      "loss": 0.3234,
      "step": 1330
    },
    {
      "epoch": 1.1343915343915345,
      "grad_norm": 1.1562023839114728,
      "learning_rate": 3.950469448508532e-05,
      "loss": 0.3143,
      "step": 1340
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 1.1140077209285717,
      "learning_rate": 3.885542969347472e-05,
      "loss": 0.2943,
      "step": 1350
    },
    {
      "epoch": 1.1513227513227513,
      "grad_norm": 1.3253628791792769,
      "learning_rate": 3.8208136402333314e-05,
      "loss": 0.3452,
      "step": 1360
    },
    {
      "epoch": 1.1597883597883598,
      "grad_norm": 1.1124786786593033,
      "learning_rate": 3.756292911935339e-05,
      "loss": 0.3371,
      "step": 1370
    },
    {
      "epoch": 1.1682539682539683,
      "grad_norm": 1.199051505503681,
      "learning_rate": 3.691992198320753e-05,
      "loss": 0.3602,
      "step": 1380
    },
    {
      "epoch": 1.1767195767195768,
      "grad_norm": 1.1816565721598766,
      "learning_rate": 3.627922874335716e-05,
      "loss": 0.3441,
      "step": 1390
    },
    {
      "epoch": 1.1851851851851851,
      "grad_norm": 1.1528028837948154,
      "learning_rate": 3.564096273993012e-05,
      "loss": 0.3354,
      "step": 1400
    },
    {
      "epoch": 1.1936507936507936,
      "grad_norm": 1.1065642388382553,
      "learning_rate": 3.5005236883670436e-05,
      "loss": 0.3161,
      "step": 1410
    },
    {
      "epoch": 1.2021164021164021,
      "grad_norm": 1.2712527045188007,
      "learning_rate": 3.437216363596418e-05,
      "loss": 0.3304,
      "step": 1420
    },
    {
      "epoch": 1.2105820105820106,
      "grad_norm": 1.14541451693123,
      "learning_rate": 3.37418549889448e-05,
      "loss": 0.3302,
      "step": 1430
    },
    {
      "epoch": 1.2190476190476192,
      "grad_norm": 1.0859928515093296,
      "learning_rate": 3.3114422445681425e-05,
      "loss": 0.3107,
      "step": 1440
    },
    {
      "epoch": 1.2275132275132274,
      "grad_norm": 1.2559757962594507,
      "learning_rate": 3.2489977000453745e-05,
      "loss": 0.3424,
      "step": 1450
    },
    {
      "epoch": 1.235978835978836,
      "grad_norm": 1.1302399638013196,
      "learning_rate": 3.1868629119116794e-05,
      "loss": 0.3225,
      "step": 1460
    },
    {
      "epoch": 1.2444444444444445,
      "grad_norm": 1.182346821657365,
      "learning_rate": 3.1250488719559383e-05,
      "loss": 0.3232,
      "step": 1470
    },
    {
      "epoch": 1.252910052910053,
      "grad_norm": 1.2327829899094018,
      "learning_rate": 3.0635665152259295e-05,
      "loss": 0.3288,
      "step": 1480
    },
    {
      "epoch": 1.2613756613756615,
      "grad_norm": 1.1858722427961588,
      "learning_rate": 3.002426718093897e-05,
      "loss": 0.3309,
      "step": 1490
    },
    {
      "epoch": 1.2698412698412698,
      "grad_norm": 1.0878551421788079,
      "learning_rate": 2.9416402963324896e-05,
      "loss": 0.3339,
      "step": 1500
    },
    {
      "epoch": 1.2783068783068783,
      "grad_norm": 1.3522545182756271,
      "learning_rate": 2.881218003201437e-05,
      "loss": 0.3191,
      "step": 1510
    },
    {
      "epoch": 1.2867724867724868,
      "grad_norm": 1.1604905763699422,
      "learning_rate": 2.821170527545254e-05,
      "loss": 0.3191,
      "step": 1520
    },
    {
      "epoch": 1.2952380952380953,
      "grad_norm": 1.0896975112303264,
      "learning_rate": 2.7615084919023794e-05,
      "loss": 0.3278,
      "step": 1530
    },
    {
      "epoch": 1.3037037037037038,
      "grad_norm": 1.2783898410901267,
      "learning_rate": 2.7022424506260113e-05,
      "loss": 0.3552,
      "step": 1540
    },
    {
      "epoch": 1.312169312169312,
      "grad_norm": 1.2412629592249576,
      "learning_rate": 2.643382888017022e-05,
      "loss": 0.3177,
      "step": 1550
    },
    {
      "epoch": 1.3206349206349206,
      "grad_norm": 1.182083170394083,
      "learning_rate": 2.5849402164692593e-05,
      "loss": 0.352,
      "step": 1560
    },
    {
      "epoch": 1.3291005291005291,
      "grad_norm": 1.2384377434425395,
      "learning_rate": 2.5269247746275716e-05,
      "loss": 0.3186,
      "step": 1570
    },
    {
      "epoch": 1.3375661375661376,
      "grad_norm": 1.1317133426240273,
      "learning_rate": 2.4693468255588752e-05,
      "loss": 0.3143,
      "step": 1580
    },
    {
      "epoch": 1.3460317460317461,
      "grad_norm": 1.127659032580244,
      "learning_rate": 2.412216554936595e-05,
      "loss": 0.3043,
      "step": 1590
    },
    {
      "epoch": 1.3544973544973544,
      "grad_norm": 1.1270117058544764,
      "learning_rate": 2.3555440692387947e-05,
      "loss": 0.3122,
      "step": 1600
    },
    {
      "epoch": 1.362962962962963,
      "grad_norm": 1.1012994404894896,
      "learning_rate": 2.2993393939603186e-05,
      "loss": 0.2964,
      "step": 1610
    },
    {
      "epoch": 1.3714285714285714,
      "grad_norm": 1.0779235542307226,
      "learning_rate": 2.2436124718392565e-05,
      "loss": 0.3206,
      "step": 1620
    },
    {
      "epoch": 1.37989417989418,
      "grad_norm": 1.169194411783028,
      "learning_rate": 2.1883731610980518e-05,
      "loss": 0.3057,
      "step": 1630
    },
    {
      "epoch": 1.3883597883597885,
      "grad_norm": 1.073218506562628,
      "learning_rate": 2.1336312336995572e-05,
      "loss": 0.3196,
      "step": 1640
    },
    {
      "epoch": 1.3968253968253967,
      "grad_norm": 1.1448212371915947,
      "learning_rate": 2.0793963736183507e-05,
      "loss": 0.3023,
      "step": 1650
    },
    {
      "epoch": 1.4052910052910053,
      "grad_norm": 1.3562963055025201,
      "learning_rate": 2.025678175127623e-05,
      "loss": 0.3344,
      "step": 1660
    },
    {
      "epoch": 1.4137566137566138,
      "grad_norm": 1.2735177478315347,
      "learning_rate": 1.972486141101922e-05,
      "loss": 0.3224,
      "step": 1670
    },
    {
      "epoch": 1.4222222222222223,
      "grad_norm": 1.1826426943582573,
      "learning_rate": 1.9198296813360777e-05,
      "loss": 0.3279,
      "step": 1680
    },
    {
      "epoch": 1.4306878306878308,
      "grad_norm": 1.1291458767171278,
      "learning_rate": 1.8677181108805823e-05,
      "loss": 0.3126,
      "step": 1690
    },
    {
      "epoch": 1.439153439153439,
      "grad_norm": 1.2660744922449119,
      "learning_rate": 1.8161606483937433e-05,
      "loss": 0.3335,
      "step": 1700
    },
    {
      "epoch": 1.4476190476190476,
      "grad_norm": 1.0518342184370524,
      "learning_rate": 1.7651664145108725e-05,
      "loss": 0.3149,
      "step": 1710
    },
    {
      "epoch": 1.456084656084656,
      "grad_norm": 1.2016784882503588,
      "learning_rate": 1.7147444302308273e-05,
      "loss": 0.3243,
      "step": 1720
    },
    {
      "epoch": 1.4645502645502646,
      "grad_norm": 1.1815033136893196,
      "learning_rate": 1.6649036153201797e-05,
      "loss": 0.3219,
      "step": 1730
    },
    {
      "epoch": 1.4730158730158731,
      "grad_norm": 1.1655388711655206,
      "learning_rate": 1.6156527867352845e-05,
      "loss": 0.3187,
      "step": 1740
    },
    {
      "epoch": 1.4814814814814814,
      "grad_norm": 1.2317123234341782,
      "learning_rate": 1.5670006570625396e-05,
      "loss": 0.2999,
      "step": 1750
    },
    {
      "epoch": 1.48994708994709,
      "grad_norm": 1.2768357746714913,
      "learning_rate": 1.5189558329771092e-05,
      "loss": 0.3237,
      "step": 1760
    },
    {
      "epoch": 1.4984126984126984,
      "grad_norm": 1.1451972982515424,
      "learning_rate": 1.4715268137203852e-05,
      "loss": 0.296,
      "step": 1770
    },
    {
      "epoch": 1.506878306878307,
      "grad_norm": 1.2388164845125278,
      "learning_rate": 1.4247219895964458e-05,
      "loss": 0.337,
      "step": 1780
    },
    {
      "epoch": 1.5153439153439154,
      "grad_norm": 1.196613109479644,
      "learning_rate": 1.3785496404877967e-05,
      "loss": 0.3117,
      "step": 1790
    },
    {
      "epoch": 1.5238095238095237,
      "grad_norm": 1.0738735120315672,
      "learning_rate": 1.3330179343906424e-05,
      "loss": 0.3236,
      "step": 1800
    },
    {
      "epoch": 1.5322751322751322,
      "grad_norm": 1.1319149412117087,
      "learning_rate": 1.2881349259699416e-05,
      "loss": 0.309,
      "step": 1810
    },
    {
      "epoch": 1.5407407407407407,
      "grad_norm": 1.1664826422000727,
      "learning_rate": 1.2439085551345209e-05,
      "loss": 0.3251,
      "step": 1820
    },
    {
      "epoch": 1.5492063492063493,
      "grad_norm": 1.1591594193359072,
      "learning_rate": 1.2003466456324907e-05,
      "loss": 0.2985,
      "step": 1830
    },
    {
      "epoch": 1.5576719576719578,
      "grad_norm": 1.2020574172279288,
      "learning_rate": 1.1574569036671978e-05,
      "loss": 0.3269,
      "step": 1840
    },
    {
      "epoch": 1.566137566137566,
      "grad_norm": 1.1386243506460356,
      "learning_rate": 1.1152469165339885e-05,
      "loss": 0.305,
      "step": 1850
    },
    {
      "epoch": 1.5746031746031746,
      "grad_norm": 1.085874089650789,
      "learning_rate": 1.0737241512779928e-05,
      "loss": 0.3173,
      "step": 1860
    },
    {
      "epoch": 1.583068783068783,
      "grad_norm": 1.197846294368878,
      "learning_rate": 1.0328959533731886e-05,
      "loss": 0.3167,
      "step": 1870
    },
    {
      "epoch": 1.5915343915343916,
      "grad_norm": 1.1368052438876846,
      "learning_rate": 9.927695454229713e-06,
      "loss": 0.2983,
      "step": 1880
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.121353872740907,
      "learning_rate": 9.533520258824557e-06,
      "loss": 0.3074,
      "step": 1890
    },
    {
      "epoch": 1.6084656084656084,
      "grad_norm": 1.3293469389715453,
      "learning_rate": 9.146503678027451e-06,
      "loss": 0.312,
      "step": 1900
    },
    {
      "epoch": 1.6169312169312169,
      "grad_norm": 1.1026868387450763,
      "learning_rate": 8.766714175973817e-06,
      "loss": 0.2977,
      "step": 1910
    },
    {
      "epoch": 1.6253968253968254,
      "grad_norm": 1.1989702128623307,
      "learning_rate": 8.394218938311982e-06,
      "loss": 0.3064,
      "step": 1920
    },
    {
      "epoch": 1.633862433862434,
      "grad_norm": 1.2736653421212538,
      "learning_rate": 8.029083860317915e-06,
      "loss": 0.3128,
      "step": 1930
    },
    {
      "epoch": 1.6423280423280424,
      "grad_norm": 1.1811892666540489,
      "learning_rate": 7.671373535238168e-06,
      "loss": 0.3331,
      "step": 1940
    },
    {
      "epoch": 1.6507936507936507,
      "grad_norm": 1.251659796122848,
      "learning_rate": 7.3211512428632115e-06,
      "loss": 0.3035,
      "step": 1950
    },
    {
      "epoch": 1.6592592592592592,
      "grad_norm": 1.0585624571230603,
      "learning_rate": 6.978478938333055e-06,
      "loss": 0.2911,
      "step": 1960
    },
    {
      "epoch": 1.6677248677248677,
      "grad_norm": 1.0285014490372995,
      "learning_rate": 6.643417241177252e-06,
      "loss": 0.299,
      "step": 1970
    },
    {
      "epoch": 1.6761904761904762,
      "grad_norm": 1.2077574437861864,
      "learning_rate": 6.3160254245911825e-06,
      "loss": 0.3086,
      "step": 1980
    },
    {
      "epoch": 1.6846560846560847,
      "grad_norm": 1.1940524793802747,
      "learning_rate": 5.9963614049504625e-06,
      "loss": 0.2972,
      "step": 1990
    },
    {
      "epoch": 1.693121693121693,
      "grad_norm": 1.2956482047870082,
      "learning_rate": 5.684481731565433e-06,
      "loss": 0.2978,
      "step": 2000
    },
    {
      "epoch": 1.7015873015873015,
      "grad_norm": 1.07352837871318,
      "learning_rate": 5.380441576677409e-06,
      "loss": 0.302,
      "step": 2010
    },
    {
      "epoch": 1.71005291005291,
      "grad_norm": 1.1429522648599941,
      "learning_rate": 5.0842947256986014e-06,
      "loss": 0.3171,
      "step": 2020
    },
    {
      "epoch": 1.7185185185185186,
      "grad_norm": 1.2683501315969479,
      "learning_rate": 4.796093567697413e-06,
      "loss": 0.3059,
      "step": 2030
    },
    {
      "epoch": 1.726984126984127,
      "grad_norm": 1.134598484064839,
      "learning_rate": 4.515889086130565e-06,
      "loss": 0.311,
      "step": 2040
    },
    {
      "epoch": 1.7354497354497354,
      "grad_norm": 1.2457646181888378,
      "learning_rate": 4.243730849824101e-06,
      "loss": 0.2849,
      "step": 2050
    },
    {
      "epoch": 1.7439153439153439,
      "grad_norm": 1.2046362639691561,
      "learning_rate": 3.979667004204535e-06,
      "loss": 0.2962,
      "step": 2060
    },
    {
      "epoch": 1.7523809523809524,
      "grad_norm": 1.1899110108249271,
      "learning_rate": 3.7237442627817753e-06,
      "loss": 0.3174,
      "step": 2070
    },
    {
      "epoch": 1.7608465608465609,
      "grad_norm": 1.1957979006872952,
      "learning_rate": 3.4760078988854305e-06,
      "loss": 0.2868,
      "step": 2080
    },
    {
      "epoch": 1.7693121693121694,
      "grad_norm": 1.0992839797811276,
      "learning_rate": 3.236501737655845e-06,
      "loss": 0.3062,
      "step": 2090
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 1.2094059824739993,
      "learning_rate": 3.005268148291296e-06,
      "loss": 0.3024,
      "step": 2100
    },
    {
      "epoch": 1.7862433862433862,
      "grad_norm": 1.2877529846662792,
      "learning_rate": 2.7823480365528345e-06,
      "loss": 0.3109,
      "step": 2110
    },
    {
      "epoch": 1.7947089947089947,
      "grad_norm": 1.0626267084104033,
      "learning_rate": 2.5677808375279443e-06,
      "loss": 0.2933,
      "step": 2120
    },
    {
      "epoch": 1.8031746031746032,
      "grad_norm": 1.2728324345583901,
      "learning_rate": 2.3616045086543637e-06,
      "loss": 0.313,
      "step": 2130
    },
    {
      "epoch": 1.8116402116402117,
      "grad_norm": 1.1679924754572557,
      "learning_rate": 2.1638555230053315e-06,
      "loss": 0.3177,
      "step": 2140
    },
    {
      "epoch": 1.82010582010582,
      "grad_norm": 1.2780416859602457,
      "learning_rate": 1.974568862837445e-06,
      "loss": 0.3161,
      "step": 2150
    },
    {
      "epoch": 1.8285714285714287,
      "grad_norm": 1.1373367274657828,
      "learning_rate": 1.7937780134021808e-06,
      "loss": 0.3064,
      "step": 2160
    },
    {
      "epoch": 1.837037037037037,
      "grad_norm": 1.371545482098203,
      "learning_rate": 1.62151495702228e-06,
      "loss": 0.2827,
      "step": 2170
    },
    {
      "epoch": 1.8455026455026455,
      "grad_norm": 1.2124638052571406,
      "learning_rate": 1.4578101674340372e-06,
      "loss": 0.2928,
      "step": 2180
    },
    {
      "epoch": 1.853968253968254,
      "grad_norm": 1.3282762207991974,
      "learning_rate": 1.3026926043963717e-06,
      "loss": 0.312,
      "step": 2190
    },
    {
      "epoch": 1.8624338624338623,
      "grad_norm": 1.2169277319430767,
      "learning_rate": 1.1561897085678341e-06,
      "loss": 0.3267,
      "step": 2200
    },
    {
      "epoch": 1.870899470899471,
      "grad_norm": 1.283700664779273,
      "learning_rate": 1.0183273966522356e-06,
      "loss": 0.302,
      "step": 2210
    },
    {
      "epoch": 1.8793650793650793,
      "grad_norm": 1.1764816717143525,
      "learning_rate": 8.891300568139638e-07,
      "loss": 0.2905,
      "step": 2220
    },
    {
      "epoch": 1.8878306878306879,
      "grad_norm": 1.1049909165614302,
      "learning_rate": 7.686205443636407e-07,
      "loss": 0.3135,
      "step": 2230
    },
    {
      "epoch": 1.8962962962962964,
      "grad_norm": 1.1865755254945776,
      "learning_rate": 6.568201777149729e-07,
      "loss": 0.2974,
      "step": 2240
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 1.2726871875485801,
      "learning_rate": 5.537487346134629e-07,
      "loss": 0.2736,
      "step": 2250
    },
    {
      "epoch": 1.9132275132275134,
      "grad_norm": 1.1770066603089386,
      "learning_rate": 4.594244486376964e-07,
      "loss": 0.3008,
      "step": 2260
    },
    {
      "epoch": 1.9216931216931217,
      "grad_norm": 1.1745505888575511,
      "learning_rate": 3.738640059737619e-07,
      "loss": 0.317,
      "step": 2270
    },
    {
      "epoch": 1.9301587301587302,
      "grad_norm": 1.2261734009974063,
      "learning_rate": 2.9708254246344494e-07,
      "loss": 0.295,
      "step": 2280
    },
    {
      "epoch": 1.9386243386243387,
      "grad_norm": 1.2436623050786833,
      "learning_rate": 2.2909364092663643e-07,
      "loss": 0.3023,
      "step": 2290
    },
    {
      "epoch": 1.947089947089947,
      "grad_norm": 1.2153390698718014,
      "learning_rate": 1.699093287585274e-07,
      "loss": 0.3012,
      "step": 2300
    },
    {
      "epoch": 1.9555555555555557,
      "grad_norm": 1.21126029855381,
      "learning_rate": 1.1954007580192718e-07,
      "loss": 0.3165,
      "step": 2310
    },
    {
      "epoch": 1.964021164021164,
      "grad_norm": 1.1700440086578683,
      "learning_rate": 7.799479249510633e-08,
      "loss": 0.3153,
      "step": 2320
    },
    {
      "epoch": 1.9724867724867725,
      "grad_norm": 1.2008513089503834,
      "learning_rate": 4.528082829553526e-08,
      "loss": 0.2764,
      "step": 2330
    },
    {
      "epoch": 1.980952380952381,
      "grad_norm": 1.096133202492879,
      "learning_rate": 2.14039703797575e-08,
      "loss": 0.3095,
      "step": 2340
    },
    {
      "epoch": 1.9894179894179893,
      "grad_norm": 1.1416782493668638,
      "learning_rate": 6.36844261959757e-09,
      "loss": 0.2999,
      "step": 2350
    },
    {
      "epoch": 1.997883597883598,
      "grad_norm": 1.2624410334213494,
      "learning_rate": 1.7690483496979326e-10,
      "loss": 0.3005,
      "step": 2360
    },
    {
      "epoch": 1.9995767195767196,
      "step": 2362,
      "total_flos": 434751562907648.0,
      "train_loss": 0.3921900775147938,
      "train_runtime": 21626.3872,
      "train_samples_per_second": 0.874,
      "train_steps_per_second": 0.109
    }
  ],
  "logging_steps": 10,
  "max_steps": 2362,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 434751562907648.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}