{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.3437607425232039,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001145869141744013,
      "grad_norm": 0.4403824210166931,
      "learning_rate": 3e-06,
      "loss": 2.2215,
      "step": 10
    },
    {
      "epoch": 0.002291738283488026,
      "grad_norm": 0.3687293827533722,
      "learning_rate": 6.333333333333334e-06,
      "loss": 2.3916,
      "step": 20
    },
    {
      "epoch": 0.0034376074252320385,
      "grad_norm": 0.2981606125831604,
      "learning_rate": 9.666666666666667e-06,
      "loss": 1.9973,
      "step": 30
    },
    {
      "epoch": 0.004583476566976052,
      "grad_norm": 0.33264994621276855,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 2.1194,
      "step": 40
    },
    {
      "epoch": 0.005729345708720064,
      "grad_norm": 0.30116188526153564,
      "learning_rate": 1.6333333333333335e-05,
      "loss": 1.8737,
      "step": 50
    },
    {
      "epoch": 0.006875214850464077,
      "grad_norm": 0.2202555388212204,
      "learning_rate": 1.9666666666666666e-05,
      "loss": 1.997,
      "step": 60
    },
    {
      "epoch": 0.00802108399220809,
      "grad_norm": 0.3031713366508484,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 1.7876,
      "step": 70
    },
    {
      "epoch": 0.009166953133952103,
      "grad_norm": 0.2614346444606781,
      "learning_rate": 2.633333333333333e-05,
      "loss": 1.8431,
      "step": 80
    },
    {
      "epoch": 0.010312822275696116,
      "grad_norm": 0.5091195106506348,
      "learning_rate": 2.9666666666666672e-05,
      "loss": 1.7054,
      "step": 90
    },
    {
      "epoch": 0.011458691417440128,
      "grad_norm": 0.48259875178337097,
      "learning_rate": 3.3e-05,
      "loss": 1.6286,
      "step": 100
    },
    {
      "epoch": 0.01260456055918414,
      "grad_norm": 0.27754926681518555,
      "learning_rate": 3.633333333333333e-05,
      "loss": 1.7845,
      "step": 110
    },
    {
      "epoch": 0.013750429700928154,
      "grad_norm": 0.40767142176628113,
      "learning_rate": 3.966666666666667e-05,
      "loss": 1.6452,
      "step": 120
    },
    {
      "epoch": 0.014896298842672167,
      "grad_norm": 0.3088354170322418,
      "learning_rate": 4.3e-05,
      "loss": 1.8875,
      "step": 130
    },
    {
      "epoch": 0.01604216798441618,
      "grad_norm": 0.4631209671497345,
      "learning_rate": 4.633333333333333e-05,
      "loss": 1.664,
      "step": 140
    },
    {
      "epoch": 0.017188037126160193,
      "grad_norm": 0.45035237073898315,
      "learning_rate": 4.966666666666667e-05,
      "loss": 1.8095,
      "step": 150
    },
    {
      "epoch": 0.018333906267904206,
      "grad_norm": 0.25533923506736755,
      "learning_rate": 5.300000000000001e-05,
      "loss": 1.7789,
      "step": 160
    },
    {
      "epoch": 0.01947977540964822,
      "grad_norm": 0.46780475974082947,
      "learning_rate": 5.633333333333334e-05,
      "loss": 1.7433,
      "step": 170
    },
    {
      "epoch": 0.020625644551392232,
      "grad_norm": 0.5887249112129211,
      "learning_rate": 5.966666666666667e-05,
      "loss": 1.6182,
      "step": 180
    },
    {
      "epoch": 0.021771513693136242,
      "grad_norm": 0.4282844066619873,
      "learning_rate": 6.3e-05,
      "loss": 1.7491,
      "step": 190
    },
    {
      "epoch": 0.022917382834880255,
      "grad_norm": 0.4257371723651886,
      "learning_rate": 6.633333333333334e-05,
      "loss": 1.66,
      "step": 200
    },
    {
      "epoch": 0.02406325197662427,
      "grad_norm": 0.35298994183540344,
      "learning_rate": 6.966666666666668e-05,
      "loss": 1.6686,
      "step": 210
    },
    {
      "epoch": 0.02520912111836828,
      "grad_norm": 0.3789626657962799,
      "learning_rate": 7.3e-05,
      "loss": 1.778,
      "step": 220
    },
    {
      "epoch": 0.026354990260112295,
      "grad_norm": 0.34714069962501526,
      "learning_rate": 7.633333333333334e-05,
      "loss": 1.6792,
      "step": 230
    },
    {
      "epoch": 0.027500859401856308,
      "grad_norm": 0.283817857503891,
      "learning_rate": 7.966666666666666e-05,
      "loss": 1.652,
      "step": 240
    },
    {
      "epoch": 0.02864672854360032,
      "grad_norm": 0.524991512298584,
      "learning_rate": 8.3e-05,
      "loss": 1.6887,
      "step": 250
    },
    {
      "epoch": 0.029792597685344334,
      "grad_norm": 0.39225083589553833,
      "learning_rate": 8.633333333333334e-05,
      "loss": 1.82,
      "step": 260
    },
    {
      "epoch": 0.030938466827088347,
      "grad_norm": 0.517483115196228,
      "learning_rate": 8.966666666666666e-05,
      "loss": 1.5324,
      "step": 270
    },
    {
      "epoch": 0.03208433596883236,
      "grad_norm": 0.30848997831344604,
      "learning_rate": 9.300000000000001e-05,
      "loss": 1.6838,
      "step": 280
    },
    {
      "epoch": 0.03323020511057637,
      "grad_norm": 0.4000117778778076,
      "learning_rate": 9.633333333333335e-05,
      "loss": 1.5693,
      "step": 290
    },
    {
      "epoch": 0.034376074252320386,
      "grad_norm": 0.34917208552360535,
      "learning_rate": 9.966666666666667e-05,
      "loss": 1.6552,
      "step": 300
    },
    {
      "epoch": 0.0355219433940644,
      "grad_norm": 0.375290185213089,
      "learning_rate": 9.999725846827562e-05,
      "loss": 1.7485,
      "step": 310
    },
    {
      "epoch": 0.03666781253580841,
      "grad_norm": 0.31680381298065186,
      "learning_rate": 9.998778195446311e-05,
      "loss": 1.4147,
      "step": 320
    },
    {
      "epoch": 0.037813681677552426,
      "grad_norm": 0.43682149052619934,
      "learning_rate": 9.997153789515461e-05,
      "loss": 1.6653,
      "step": 330
    },
    {
      "epoch": 0.03895955081929644,
      "grad_norm": 0.3149017095565796,
      "learning_rate": 9.994852848953574e-05,
      "loss": 1.6447,
      "step": 340
    },
    {
      "epoch": 0.04010541996104045,
      "grad_norm": 0.5677515268325806,
      "learning_rate": 9.991875685271168e-05,
      "loss": 1.5702,
      "step": 350
    },
    {
      "epoch": 0.041251289102784465,
      "grad_norm": 0.5448745489120483,
      "learning_rate": 9.988222701528547e-05,
      "loss": 1.6931,
      "step": 360
    },
    {
      "epoch": 0.04239715824452848,
      "grad_norm": 0.3963087201118469,
      "learning_rate": 9.983894392281237e-05,
      "loss": 1.7516,
      "step": 370
    },
    {
      "epoch": 0.043543027386272484,
      "grad_norm": 0.43567872047424316,
      "learning_rate": 9.978891343513023e-05,
      "loss": 1.6208,
      "step": 380
    },
    {
      "epoch": 0.0446888965280165,
      "grad_norm": 0.5545846819877625,
      "learning_rate": 9.973214232556622e-05,
      "loss": 1.6865,
      "step": 390
    },
    {
      "epoch": 0.04583476566976051,
      "grad_norm": 0.4242377281188965,
      "learning_rate": 9.966863828001982e-05,
      "loss": 1.549,
      "step": 400
    },
    {
      "epoch": 0.046980634811504524,
      "grad_norm": 0.447827011346817,
      "learning_rate": 9.959840989592226e-05,
      "loss": 1.4388,
      "step": 410
    },
    {
      "epoch": 0.04812650395324854,
      "grad_norm": 0.37601059675216675,
      "learning_rate": 9.952146668107254e-05,
      "loss": 1.4154,
      "step": 420
    },
    {
      "epoch": 0.04927237309499255,
      "grad_norm": 0.5751256346702576,
      "learning_rate": 9.94378190523503e-05,
      "loss": 1.4273,
      "step": 430
    },
    {
      "epoch": 0.05041824223673656,
      "grad_norm": 0.508073091506958,
      "learning_rate": 9.934747833430547e-05,
      "loss": 1.5258,
      "step": 440
    },
    {
      "epoch": 0.051564111378480576,
      "grad_norm": 0.4049976170063019,
      "learning_rate": 9.925045675762514e-05,
      "loss": 1.7881,
      "step": 450
    },
    {
      "epoch": 0.05270998052022459,
      "grad_norm": 0.483882337808609,
      "learning_rate": 9.914676745747772e-05,
      "loss": 1.6563,
      "step": 460
    },
    {
      "epoch": 0.0538558496619686,
      "grad_norm": 0.44065427780151367,
      "learning_rate": 9.903642447173465e-05,
      "loss": 1.4789,
      "step": 470
    },
    {
      "epoch": 0.055001718803712615,
      "grad_norm": 0.43570518493652344,
      "learning_rate": 9.891944273906986e-05,
      "loss": 1.5056,
      "step": 480
    },
    {
      "epoch": 0.05614758794545663,
      "grad_norm": 0.5358373522758484,
      "learning_rate": 9.879583809693738e-05,
      "loss": 1.3754,
      "step": 490
    },
    {
      "epoch": 0.05729345708720064,
      "grad_norm": 0.5095910429954529,
      "learning_rate": 9.866562727942714e-05,
      "loss": 1.624,
      "step": 500
    },
    {
      "epoch": 0.058439326228944655,
      "grad_norm": 0.6042481064796448,
      "learning_rate": 9.85288279149995e-05,
      "loss": 1.6504,
      "step": 510
    },
    {
      "epoch": 0.05958519537068867,
      "grad_norm": 0.5279871821403503,
      "learning_rate": 9.838545852409857e-05,
      "loss": 1.6803,
      "step": 520
    },
    {
      "epoch": 0.06073106451243268,
      "grad_norm": 0.34437739849090576,
      "learning_rate": 9.823553851664489e-05,
      "loss": 1.4585,
      "step": 530
    },
    {
      "epoch": 0.061876933654176694,
      "grad_norm": 0.4585515856742859,
      "learning_rate": 9.807908818940761e-05,
      "loss": 1.6015,
      "step": 540
    },
    {
      "epoch": 0.06302280279592071,
      "grad_norm": 0.43472760915756226,
      "learning_rate": 9.791612872325667e-05,
      "loss": 1.6088,
      "step": 550
    },
    {
      "epoch": 0.06416867193766472,
      "grad_norm": 0.5794304609298706,
      "learning_rate": 9.77466821802952e-05,
      "loss": 1.6964,
      "step": 560
    },
    {
      "epoch": 0.06531454107940873,
      "grad_norm": 0.5335965752601624,
      "learning_rate": 9.75707715008727e-05,
      "loss": 1.4224,
      "step": 570
    },
    {
      "epoch": 0.06646041022115275,
      "grad_norm": 0.572496771812439,
      "learning_rate": 9.73884205004793e-05,
      "loss": 1.5168,
      "step": 580
    },
    {
      "epoch": 0.06760627936289676,
      "grad_norm": 0.40056779980659485,
      "learning_rate": 9.719965386652141e-05,
      "loss": 1.5301,
      "step": 590
    },
    {
      "epoch": 0.06875214850464077,
      "grad_norm": 0.43687576055526733,
      "learning_rate": 9.700449715497961e-05,
      "loss": 1.6744,
      "step": 600
    },
    {
      "epoch": 0.06989801764638479,
      "grad_norm": 0.45747336745262146,
      "learning_rate": 9.680297678694867e-05,
      "loss": 1.6185,
      "step": 610
    },
    {
      "epoch": 0.0710438867881288,
      "grad_norm": 0.40091991424560547,
      "learning_rate": 9.659512004506057e-05,
      "loss": 1.5316,
      "step": 620
    },
    {
      "epoch": 0.07218975592987281,
      "grad_norm": 0.4765830338001251,
      "learning_rate": 9.63809550697909e-05,
      "loss": 1.599,
      "step": 630
    },
    {
      "epoch": 0.07333562507161683,
      "grad_norm": 0.5011415481567383,
      "learning_rate": 9.616051085564906e-05,
      "loss": 1.7098,
      "step": 640
    },
    {
      "epoch": 0.07448149421336084,
      "grad_norm": 0.5357868671417236,
      "learning_rate": 9.593381724725285e-05,
      "loss": 1.7165,
      "step": 650
    },
    {
      "epoch": 0.07562736335510485,
      "grad_norm": 0.5434888005256653,
      "learning_rate": 9.570090493528809e-05,
      "loss": 1.622,
      "step": 660
    },
    {
      "epoch": 0.07677323249684886,
      "grad_norm": 0.39640671014785767,
      "learning_rate": 9.546180545235344e-05,
      "loss": 1.5643,
      "step": 670
    },
    {
      "epoch": 0.07791910163859288,
      "grad_norm": 0.4940226078033447,
      "learning_rate": 9.52165511686915e-05,
      "loss": 1.4481,
      "step": 680
    },
    {
      "epoch": 0.07906497078033689,
      "grad_norm": 0.6464428901672363,
      "learning_rate": 9.496517528780637e-05,
      "loss": 1.6744,
      "step": 690
    },
    {
      "epoch": 0.0802108399220809,
      "grad_norm": 0.6630286574363708,
      "learning_rate": 9.47077118419684e-05,
      "loss": 1.4707,
      "step": 700
    },
    {
      "epoch": 0.08135670906382492,
      "grad_norm": 0.564497709274292,
      "learning_rate": 9.444419568760684e-05,
      "loss": 1.571,
      "step": 710
    },
    {
      "epoch": 0.08250257820556893,
      "grad_norm": 0.8117035031318665,
      "learning_rate": 9.417466250059073e-05,
      "loss": 1.5533,
      "step": 720
    },
    {
      "epoch": 0.08364844734731294,
      "grad_norm": 0.6528947353363037,
      "learning_rate": 9.389914877139903e-05,
      "loss": 1.6647,
      "step": 730
    },
    {
      "epoch": 0.08479431648905696,
      "grad_norm": 0.5657352209091187,
      "learning_rate": 9.361769180018038e-05,
      "loss": 1.6766,
      "step": 740
    },
    {
      "epoch": 0.08594018563080097,
      "grad_norm": 0.5918785333633423,
      "learning_rate": 9.333032969170326e-05,
      "loss": 1.6842,
      "step": 750
    },
    {
      "epoch": 0.08708605477254497,
      "grad_norm": 0.5485035181045532,
      "learning_rate": 9.30371013501972e-05,
      "loss": 1.3688,
      "step": 760
    },
    {
      "epoch": 0.08823192391428898,
      "grad_norm": 0.49867671728134155,
      "learning_rate": 9.273804647408575e-05,
      "loss": 1.7568,
      "step": 770
    },
    {
      "epoch": 0.089377793056033,
      "grad_norm": 0.7494494915008545,
      "learning_rate": 9.243320555061205e-05,
      "loss": 1.4289,
      "step": 780
    },
    {
      "epoch": 0.09052366219777701,
      "grad_norm": 0.5496593713760376,
      "learning_rate": 9.212261985035739e-05,
      "loss": 1.6298,
      "step": 790
    },
    {
      "epoch": 0.09166953133952102,
      "grad_norm": 0.38885053992271423,
      "learning_rate": 9.180633142165384e-05,
      "loss": 1.3307,
      "step": 800
    },
    {
      "epoch": 0.09281540048126503,
      "grad_norm": 0.5980333685874939,
      "learning_rate": 9.148438308489168e-05,
      "loss": 1.7473,
      "step": 810
    },
    {
      "epoch": 0.09396126962300905,
      "grad_norm": 0.4313773214817047,
      "learning_rate": 9.11568184267221e-05,
      "loss": 1.6697,
      "step": 820
    },
    {
      "epoch": 0.09510713876475306,
      "grad_norm": 0.4981066882610321,
      "learning_rate": 9.082368179415632e-05,
      "loss": 1.6798,
      "step": 830
    },
    {
      "epoch": 0.09625300790649707,
      "grad_norm": 0.6887238621711731,
      "learning_rate": 9.04850182885617e-05,
      "loss": 1.4729,
      "step": 840
    },
    {
      "epoch": 0.09739887704824109,
      "grad_norm": 0.5206602811813354,
      "learning_rate": 9.014087375955573e-05,
      "loss": 1.6206,
      "step": 850
    },
    {
      "epoch": 0.0985447461899851,
      "grad_norm": 0.508270263671875,
      "learning_rate": 8.979129479879873e-05,
      "loss": 1.6445,
      "step": 860
    },
    {
      "epoch": 0.09969061533172911,
      "grad_norm": 0.5993254780769348,
      "learning_rate": 8.943632873368611e-05,
      "loss": 1.3967,
      "step": 870
    },
    {
      "epoch": 0.10083648447347313,
      "grad_norm": 0.4603252112865448,
      "learning_rate": 8.907602362094094e-05,
      "loss": 1.5337,
      "step": 880
    },
    {
      "epoch": 0.10198235361521714,
      "grad_norm": 0.6201605796813965,
      "learning_rate": 8.871042824010791e-05,
      "loss": 1.5677,
      "step": 890
    },
    {
      "epoch": 0.10312822275696115,
      "grad_norm": 0.4221727252006531,
      "learning_rate": 8.833959208694929e-05,
      "loss": 1.4596,
      "step": 900
    },
    {
      "epoch": 0.10427409189870517,
      "grad_norm": 0.47249990701675415,
      "learning_rate": 8.796356536674403e-05,
      "loss": 1.495,
      "step": 910
    },
    {
      "epoch": 0.10541996104044918,
      "grad_norm": 0.49199992418289185,
      "learning_rate": 8.758239898749085e-05,
      "loss": 1.6891,
      "step": 920
    },
    {
      "epoch": 0.10656583018219319,
      "grad_norm": 0.7329378128051758,
      "learning_rate": 8.719614455301593e-05,
      "loss": 1.6418,
      "step": 930
    },
    {
      "epoch": 0.1077116993239372,
      "grad_norm": 0.5124281644821167,
      "learning_rate": 8.680485435598673e-05,
      "loss": 1.4896,
      "step": 940
    },
    {
      "epoch": 0.10885756846568122,
      "grad_norm": 0.4630810618400574,
      "learning_rate": 8.640858137083232e-05,
      "loss": 1.6428,
      "step": 950
    },
    {
      "epoch": 0.11000343760742523,
      "grad_norm": 0.5222834944725037,
      "learning_rate": 8.600737924657156e-05,
      "loss": 1.6116,
      "step": 960
    },
    {
      "epoch": 0.11114930674916924,
      "grad_norm": 0.5838509202003479,
      "learning_rate": 8.560130229954984e-05,
      "loss": 1.4356,
      "step": 970
    },
    {
      "epoch": 0.11229517589091326,
      "grad_norm": 0.486001193523407,
      "learning_rate": 8.519040550608546e-05,
      "loss": 1.4285,
      "step": 980
    },
    {
      "epoch": 0.11344104503265727,
      "grad_norm": 0.5732933282852173,
      "learning_rate": 8.477474449502682e-05,
      "loss": 1.5028,
      "step": 990
    },
    {
      "epoch": 0.11458691417440128,
      "grad_norm": 0.6634925603866577,
      "learning_rate": 8.435437554022115e-05,
      "loss": 1.4051,
      "step": 1000
    },
    {
      "epoch": 0.1157327833161453,
      "grad_norm": 0.6763463020324707,
      "learning_rate": 8.392935555289584e-05,
      "loss": 1.4876,
      "step": 1010
    },
    {
      "epoch": 0.11687865245788931,
      "grad_norm": 0.5878898501396179,
      "learning_rate": 8.349974207395366e-05,
      "loss": 1.5388,
      "step": 1020
    },
    {
      "epoch": 0.11802452159963332,
      "grad_norm": 0.5946276187896729,
      "learning_rate": 8.306559326618259e-05,
      "loss": 1.5264,
      "step": 1030
    },
    {
      "epoch": 0.11917039074137734,
      "grad_norm": 0.5235640406608582,
      "learning_rate": 8.26269679063816e-05,
      "loss": 1.6672,
      "step": 1040
    },
    {
      "epoch": 0.12031625988312135,
      "grad_norm": 0.6421871185302734,
      "learning_rate": 8.218392537740305e-05,
      "loss": 1.6887,
      "step": 1050
    },
    {
      "epoch": 0.12146212902486536,
      "grad_norm": 0.5750526189804077,
      "learning_rate": 8.173652566011338e-05,
      "loss": 1.6317,
      "step": 1060
    },
    {
      "epoch": 0.12260799816660937,
      "grad_norm": 0.755728006362915,
      "learning_rate": 8.128482932527255e-05,
      "loss": 1.5624,
      "step": 1070
    },
    {
      "epoch": 0.12375386730835339,
      "grad_norm": 0.5246460437774658,
      "learning_rate": 8.082889752533375e-05,
      "loss": 1.6643,
      "step": 1080
    },
    {
      "epoch": 0.1248997364500974,
      "grad_norm": 0.5184617042541504,
      "learning_rate": 8.036879198616434e-05,
      "loss": 1.8416,
      "step": 1090
    },
    {
      "epoch": 0.12604560559184141,
      "grad_norm": 0.6587857604026794,
      "learning_rate": 7.990457499868919e-05,
      "loss": 1.3842,
      "step": 1100
    },
    {
      "epoch": 0.1271914747335854,
      "grad_norm": 0.5388489365577698,
      "learning_rate": 7.943630941045744e-05,
      "loss": 1.5292,
      "step": 1110
    },
    {
      "epoch": 0.12833734387532944,
      "grad_norm": 0.5552734732627869,
      "learning_rate": 7.896405861713394e-05,
      "loss": 1.322,
      "step": 1120
    },
    {
      "epoch": 0.12948321301707344,
      "grad_norm": 0.4926222264766693,
      "learning_rate": 7.848788655391658e-05,
      "loss": 1.3671,
      "step": 1130
    },
    {
      "epoch": 0.13062908215881747,
      "grad_norm": 0.49568048119544983,
      "learning_rate": 7.800785768688035e-05,
      "loss": 1.6613,
      "step": 1140
    },
    {
      "epoch": 0.13177495130056147,
      "grad_norm": 0.6832533478736877,
      "learning_rate": 7.752403700424979e-05,
      "loss": 1.6197,
      "step": 1150
    },
    {
      "epoch": 0.1329208204423055,
      "grad_norm": 0.6604668498039246,
      "learning_rate": 7.703649000760053e-05,
      "loss": 1.5766,
      "step": 1160
    },
    {
      "epoch": 0.1340666895840495,
      "grad_norm": 0.6074301600456238,
      "learning_rate": 7.654528270299154e-05,
      "loss": 1.4706,
      "step": 1170
    },
    {
      "epoch": 0.13521255872579352,
      "grad_norm": 0.6452466249465942,
      "learning_rate": 7.605048159202883e-05,
      "loss": 1.5349,
      "step": 1180
    },
    {
      "epoch": 0.13635842786753752,
      "grad_norm": 0.6840918660163879,
      "learning_rate": 7.555215366286227e-05,
      "loss": 1.4348,
      "step": 1190
    },
    {
      "epoch": 0.13750429700928155,
      "grad_norm": 0.7689867615699768,
      "learning_rate": 7.505036638111648e-05,
      "loss": 1.5708,
      "step": 1200
    },
    {
      "epoch": 0.13865016615102554,
      "grad_norm": 0.6641436815261841,
      "learning_rate": 7.454518768075704e-05,
      "loss": 1.6616,
      "step": 1210
    },
    {
      "epoch": 0.13979603529276957,
      "grad_norm": 0.5971699953079224,
      "learning_rate": 7.403668595489333e-05,
      "loss": 1.5003,
      "step": 1220
    },
    {
      "epoch": 0.14094190443451357,
      "grad_norm": 0.6612730622291565,
      "learning_rate": 7.352493004651916e-05,
      "loss": 1.4622,
      "step": 1230
    },
    {
      "epoch": 0.1420877735762576,
      "grad_norm": 0.45702728629112244,
      "learning_rate": 7.300998923919259e-05,
      "loss": 1.7201,
      "step": 1240
    },
    {
      "epoch": 0.1432336427180016,
      "grad_norm": 0.6653318405151367,
      "learning_rate": 7.249193324765599e-05,
      "loss": 1.6614,
      "step": 1250
    },
    {
      "epoch": 0.14437951185974562,
      "grad_norm": 0.4173441231250763,
      "learning_rate": 7.197083220839785e-05,
      "loss": 1.4604,
      "step": 1260
    },
    {
      "epoch": 0.14552538100148962,
      "grad_norm": 0.5349558591842651,
      "learning_rate": 7.14467566701573e-05,
      "loss": 1.5317,
      "step": 1270
    },
    {
      "epoch": 0.14667125014323365,
      "grad_norm": 0.5395480394363403,
      "learning_rate": 7.091977758437311e-05,
      "loss": 1.3445,
      "step": 1280
    },
    {
      "epoch": 0.14781711928497765,
      "grad_norm": 0.5136141180992126,
      "learning_rate": 7.038996629557783e-05,
      "loss": 1.3838,
      "step": 1290
    },
    {
      "epoch": 0.14896298842672168,
      "grad_norm": 0.4931055009365082,
      "learning_rate": 6.985739453173903e-05,
      "loss": 1.499,
      "step": 1300
    },
    {
      "epoch": 0.15010885756846568,
      "grad_norm": 0.558573305606842,
      "learning_rate": 6.932213439454837e-05,
      "loss": 1.4838,
      "step": 1310
    },
    {
      "epoch": 0.1512547267102097,
      "grad_norm": 0.6223447918891907,
      "learning_rate": 6.87842583496602e-05,
      "loss": 1.6392,
      "step": 1320
    },
    {
      "epoch": 0.1524005958519537,
      "grad_norm": 0.6831842064857483,
      "learning_rate": 6.824383921688098e-05,
      "loss": 1.6156,
      "step": 1330
    },
    {
      "epoch": 0.15354646499369773,
      "grad_norm": 0.5758393406867981,
      "learning_rate": 6.77009501603105e-05,
      "loss": 1.6267,
      "step": 1340
    },
    {
      "epoch": 0.15469233413544173,
      "grad_norm": 0.645879864692688,
      "learning_rate": 6.71556646784367e-05,
      "loss": 1.5072,
      "step": 1350
    },
    {
      "epoch": 0.15583820327718576,
      "grad_norm": 0.6415931582450867,
      "learning_rate": 6.660805659418516e-05,
      "loss": 1.6564,
      "step": 1360
    },
    {
      "epoch": 0.15698407241892975,
      "grad_norm": 0.5139831900596619,
      "learning_rate": 6.605820004492467e-05,
      "loss": 1.5847,
      "step": 1370
    },
    {
      "epoch": 0.15812994156067378,
      "grad_norm": 0.6076359152793884,
      "learning_rate": 6.550616947243009e-05,
      "loss": 1.552,
      "step": 1380
    },
    {
      "epoch": 0.15927581070241778,
      "grad_norm": 0.6446265578269958,
      "learning_rate": 6.495203961280434e-05,
      "loss": 1.4802,
      "step": 1390
    },
    {
      "epoch": 0.1604216798441618,
      "grad_norm": 0.5656043887138367,
      "learning_rate": 6.439588548636016e-05,
      "loss": 1.4842,
      "step": 1400
    },
    {
      "epoch": 0.1615675489859058,
      "grad_norm": 0.6946837902069092,
      "learning_rate": 6.38377823874636e-05,
      "loss": 1.6448,
      "step": 1410
    },
    {
      "epoch": 0.16271341812764983,
      "grad_norm": 0.5978801846504211,
      "learning_rate": 6.327780587434044e-05,
      "loss": 1.6024,
      "step": 1420
    },
    {
      "epoch": 0.16385928726939383,
      "grad_norm": 0.690424919128418,
      "learning_rate": 6.27160317588467e-05,
      "loss": 1.5692,
      "step": 1430
    },
    {
      "epoch": 0.16500515641113786,
      "grad_norm": 0.586052656173706,
      "learning_rate": 6.215253609620498e-05,
      "loss": 1.5741,
      "step": 1440
    },
    {
      "epoch": 0.16615102555288186,
      "grad_norm": 0.5687704086303711,
      "learning_rate": 6.158739517470786e-05,
      "loss": 1.4475,
      "step": 1450
    },
    {
      "epoch": 0.1672968946946259,
      "grad_norm": 0.5664721131324768,
      "learning_rate": 6.102068550538962e-05,
      "loss": 1.5458,
      "step": 1460
    },
    {
      "epoch": 0.16844276383636989,
      "grad_norm": 0.828508734703064,
      "learning_rate": 6.045248381166783e-05,
      "loss": 1.6237,
      "step": 1470
    },
    {
      "epoch": 0.1695886329781139,
      "grad_norm": 0.6727364659309387,
      "learning_rate": 5.988286701895631e-05,
      "loss": 1.6793,
      "step": 1480
    },
    {
      "epoch": 0.1707345021198579,
      "grad_norm": 0.7191620469093323,
      "learning_rate": 5.9311912244250675e-05,
      "loss": 1.7752,
      "step": 1490
    },
    {
      "epoch": 0.17188037126160194,
      "grad_norm": 0.6174139380455017,
      "learning_rate": 5.873969678568784e-05,
      "loss": 1.4516,
      "step": 1500
    },
    {
      "epoch": 0.17302624040334594,
      "grad_norm": 0.7653396129608154,
      "learning_rate": 5.816629811208112e-05,
      "loss": 1.4434,
      "step": 1510
    },
    {
      "epoch": 0.17417210954508994,
      "grad_norm": 0.5984031558036804,
      "learning_rate": 5.759179385243224e-05,
      "loss": 1.4921,
      "step": 1520
    },
    {
      "epoch": 0.17531797868683396,
      "grad_norm": 0.705712616443634,
      "learning_rate": 5.701626178542158e-05,
      "loss": 1.5121,
      "step": 1530
    },
    {
      "epoch": 0.17646384782857796,
      "grad_norm": 0.5426698327064514,
      "learning_rate": 5.643977982887815e-05,
      "loss": 1.5129,
      "step": 1540
    },
    {
      "epoch": 0.177609716970322,
      "grad_norm": 0.5686979293823242,
      "learning_rate": 5.586242602923081e-05,
      "loss": 1.5206,
      "step": 1550
    },
    {
      "epoch": 0.178755586112066,
      "grad_norm": 0.5606265664100647,
      "learning_rate": 5.528427855094206e-05,
      "loss": 1.4463,
      "step": 1560
    },
    {
      "epoch": 0.17990145525381002,
      "grad_norm": 0.5519984364509583,
      "learning_rate": 5.470541566592573e-05,
      "loss": 1.6682,
      "step": 1570
    },
    {
      "epoch": 0.18104732439555402,
      "grad_norm": 0.5270847082138062,
      "learning_rate": 5.4125915742950275e-05,
      "loss": 1.7371,
      "step": 1580
    },
    {
      "epoch": 0.18219319353729804,
      "grad_norm": 0.707064151763916,
      "learning_rate": 5.354585723702893e-05,
      "loss": 1.4248,
      "step": 1590
    },
    {
      "epoch": 0.18333906267904204,
      "grad_norm": 0.47826680541038513,
      "learning_rate": 5.296531867879809e-05,
      "loss": 1.4762,
      "step": 1600
    },
    {
      "epoch": 0.18448493182078607,
      "grad_norm": 0.64763343334198,
      "learning_rate": 5.2384378663885545e-05,
      "loss": 1.4529,
      "step": 1610
    },
    {
      "epoch": 0.18563080096253007,
      "grad_norm": 0.5846421122550964,
      "learning_rate": 5.180311584226991e-05,
      "loss": 1.4331,
      "step": 1620
    },
    {
      "epoch": 0.1867766701042741,
      "grad_norm": 0.6064422726631165,
      "learning_rate": 5.1221608907632665e-05,
      "loss": 1.5347,
      "step": 1630
    },
    {
      "epoch": 0.1879225392460181,
      "grad_norm": 0.5147915482521057,
      "learning_rate": 5.063993658670425e-05,
      "loss": 1.6031,
      "step": 1640
    },
    {
      "epoch": 0.18906840838776212,
      "grad_norm": 0.8326632380485535,
      "learning_rate": 5.0058177628605795e-05,
      "loss": 1.4734,
      "step": 1650
    },
    {
      "epoch": 0.19021427752950612,
      "grad_norm": 0.614957332611084,
      "learning_rate": 4.947641079418773e-05,
      "loss": 1.5294,
      "step": 1660
    },
    {
      "epoch": 0.19136014667125015,
      "grad_norm": 0.6000957489013672,
      "learning_rate": 4.889471484536672e-05,
      "loss": 1.4899,
      "step": 1670
    },
    {
      "epoch": 0.19250601581299415,
      "grad_norm": 0.5473623275756836,
      "learning_rate": 4.83131685344628e-05,
      "loss": 1.4639,
      "step": 1680
    },
    {
      "epoch": 0.19365188495473817,
      "grad_norm": 0.6118156313896179,
      "learning_rate": 4.773185059353732e-05,
      "loss": 1.6386,
      "step": 1690
    },
    {
      "epoch": 0.19479775409648217,
      "grad_norm": 0.9393003582954407,
      "learning_rate": 4.715083972373401e-05,
      "loss": 1.7338,
      "step": 1700
    },
    {
      "epoch": 0.1959436232382262,
      "grad_norm": 0.5811169743537903,
      "learning_rate": 4.657021458462409e-05,
      "loss": 1.4429,
      "step": 1710
    },
    {
      "epoch": 0.1970894923799702,
      "grad_norm": 0.5532368421554565,
      "learning_rate": 4.599005378355706e-05,
      "loss": 1.5809,
      "step": 1720
    },
    {
      "epoch": 0.19823536152171423,
      "grad_norm": 0.8748618960380554,
      "learning_rate": 4.541043586501842e-05,
      "loss": 1.3697,
      "step": 1730
    },
    {
      "epoch": 0.19938123066345823,
      "grad_norm": 0.8062466979026794,
      "learning_rate": 4.4831439299996084e-05,
      "loss": 1.5535,
      "step": 1740
    },
    {
      "epoch": 0.20052709980520225,
      "grad_norm": 0.9194825291633606,
      "learning_rate": 4.425314247535668e-05,
      "loss": 1.5783,
      "step": 1750
    },
    {
      "epoch": 0.20167296894694625,
      "grad_norm": 0.9836660027503967,
      "learning_rate": 4.3675623683233135e-05,
      "loss": 1.4318,
      "step": 1760
    },
    {
      "epoch": 0.20281883808869028,
      "grad_norm": 0.7427091598510742,
      "learning_rate": 4.309896111042529e-05,
      "loss": 1.7398,
      "step": 1770
    },
    {
      "epoch": 0.20396470723043428,
      "grad_norm": 0.49087536334991455,
      "learning_rate": 4.252323282781453e-05,
      "loss": 1.6288,
      "step": 1780
    },
    {
      "epoch": 0.2051105763721783,
      "grad_norm": 0.5492022037506104,
      "learning_rate": 4.1948516779794364e-05,
      "loss": 1.2943,
      "step": 1790
    },
    {
      "epoch": 0.2062564455139223,
      "grad_norm": 0.4957669675350189,
      "learning_rate": 4.137489077371787e-05,
      "loss": 1.4584,
      "step": 1800
    },
    {
      "epoch": 0.20740231465566633,
      "grad_norm": 0.631765604019165,
      "learning_rate": 4.080243246936399e-05,
      "loss": 1.6654,
      "step": 1810
    },
    {
      "epoch": 0.20854818379741033,
      "grad_norm": 0.6756720542907715,
      "learning_rate": 4.0231219368423466e-05,
      "loss": 1.402,
      "step": 1820
    },
    {
      "epoch": 0.20969405293915436,
      "grad_norm": 0.7459967732429504,
      "learning_rate": 3.9661328804006475e-05,
      "loss": 1.4162,
      "step": 1830
    },
    {
      "epoch": 0.21083992208089836,
      "grad_norm": 1.0302833318710327,
      "learning_rate": 3.9092837930172884e-05,
      "loss": 1.4953,
      "step": 1840
    },
    {
      "epoch": 0.21198579122264238,
      "grad_norm": 0.7724302411079407,
      "learning_rate": 3.852582371148687e-05,
      "loss": 1.5079,
      "step": 1850
    },
    {
      "epoch": 0.21313166036438638,
      "grad_norm": 0.5944705009460449,
      "learning_rate": 3.796036291259718e-05,
      "loss": 1.4256,
      "step": 1860
    },
    {
      "epoch": 0.2142775295061304,
      "grad_norm": 0.4870947599411011,
      "learning_rate": 3.739653208784432e-05,
      "loss": 1.679,
      "step": 1870
    },
    {
      "epoch": 0.2154233986478744,
      "grad_norm": 0.5536025166511536,
      "learning_rate": 3.683440757089646e-05,
      "loss": 1.5359,
      "step": 1880
    },
    {
      "epoch": 0.21656926778961844,
      "grad_norm": 0.6541599035263062,
      "learning_rate": 3.627406546441494e-05,
      "loss": 1.5177,
      "step": 1890
    },
    {
      "epoch": 0.21771513693136244,
      "grad_norm": 0.5040682554244995,
      "learning_rate": 3.5715581629751326e-05,
      "loss": 1.5476,
      "step": 1900
    },
    {
      "epoch": 0.21886100607310646,
      "grad_norm": 0.6246699690818787,
      "learning_rate": 3.515903167667686e-05,
      "loss": 1.4915,
      "step": 1910
    },
    {
      "epoch": 0.22000687521485046,
      "grad_norm": 0.5558972358703613,
      "learning_rate": 3.460449095314621e-05,
      "loss": 1.5111,
      "step": 1920
    },
    {
      "epoch": 0.2211527443565945,
      "grad_norm": 0.6998717784881592,
      "learning_rate": 3.40520345350965e-05,
      "loss": 1.5571,
      "step": 1930
    },
    {
      "epoch": 0.2222986134983385,
      "grad_norm": 0.6871377825737,
      "learning_rate": 3.35017372162833e-05,
      "loss": 1.4671,
      "step": 1940
    },
    {
      "epoch": 0.22344448264008251,
      "grad_norm": 0.5883384943008423,
      "learning_rate": 3.295367349815469e-05,
      "loss": 1.3532,
      "step": 1950
    },
    {
      "epoch": 0.2245903517818265,
      "grad_norm": 0.6540731191635132,
      "learning_rate": 3.240791757976491e-05,
      "loss": 1.546,
      "step": 1960
    },
    {
      "epoch": 0.22573622092357054,
      "grad_norm": 0.6483651995658875,
      "learning_rate": 3.186454334772916e-05,
      "loss": 1.5059,
      "step": 1970
    },
    {
      "epoch": 0.22688209006531454,
      "grad_norm": 0.696576714515686,
      "learning_rate": 3.132362436622035e-05,
      "loss": 1.6108,
      "step": 1980
    },
    {
      "epoch": 0.22802795920705857,
      "grad_norm": 0.610880970954895,
      "learning_rate": 3.078523386700982e-05,
      "loss": 1.5157,
      "step": 1990
    },
    {
      "epoch": 0.22917382834880257,
      "grad_norm": 0.5885791778564453,
      "learning_rate": 3.0249444739552844e-05,
      "loss": 1.5689,
      "step": 2000
    },
    {
      "epoch": 0.2303196974905466,
      "grad_norm": 0.5722992420196533,
      "learning_rate": 2.971632952112066e-05,
      "loss": 1.4526,
      "step": 2010
    },
    {
      "epoch": 0.2314655666322906,
      "grad_norm": 0.6752781271934509,
      "learning_rate": 2.918596038697995e-05,
      "loss": 1.6506,
      "step": 2020
    },
    {
      "epoch": 0.2326114357740346,
      "grad_norm": 0.5578230619430542,
      "learning_rate": 2.86584091406216e-05,
      "loss": 1.5125,
      "step": 2030
    },
    {
      "epoch": 0.23375730491577862,
      "grad_norm": 0.796444296836853,
      "learning_rate": 2.8133747204039574e-05,
      "loss": 1.3805,
      "step": 2040
    },
    {
      "epoch": 0.23490317405752262,
      "grad_norm": 0.5618960857391357,
      "learning_rate": 2.761204560806152e-05,
      "loss": 1.6502,
      "step": 2050
    },
    {
      "epoch": 0.23604904319926664,
      "grad_norm": 0.6805716156959534,
      "learning_rate": 2.709337498273243e-05,
      "loss": 1.5553,
      "step": 2060
    },
    {
      "epoch": 0.23719491234101064,
      "grad_norm": 0.5547701120376587,
      "learning_rate": 2.65778055477523e-05,
      "loss": 1.3775,
      "step": 2070
    },
    {
      "epoch": 0.23834078148275467,
      "grad_norm": 0.6155176758766174,
      "learning_rate": 2.6065407102969664e-05,
      "loss": 1.4941,
      "step": 2080
    },
    {
      "epoch": 0.23948665062449867,
      "grad_norm": 0.849638044834137,
      "learning_rate": 2.555624901893171e-05,
      "loss": 1.568,
      "step": 2090
    },
    {
      "epoch": 0.2406325197662427,
      "grad_norm": 0.5048200488090515,
      "learning_rate": 2.505040022749265e-05,
      "loss": 1.37,
      "step": 2100
    },
    {
      "epoch": 0.2417783889079867,
      "grad_norm": 0.5288580656051636,
      "learning_rate": 2.4547929212481435e-05,
      "loss": 1.3513,
      "step": 2110
    },
    {
      "epoch": 0.24292425804973072,
      "grad_norm": 0.5775280594825745,
      "learning_rate": 2.404890400043023e-05,
      "loss": 1.6431,
      "step": 2120
    },
    {
      "epoch": 0.24407012719147472,
      "grad_norm": 0.5821782350540161,
      "learning_rate": 2.3553392151364536e-05,
      "loss": 1.3928,
      "step": 2130
    },
    {
      "epoch": 0.24521599633321875,
      "grad_norm": 0.6166446805000305,
      "learning_rate": 2.3061460749656844e-05,
      "loss": 1.5129,
      "step": 2140
    },
    {
      "epoch": 0.24636186547496275,
      "grad_norm": 0.7417446374893188,
      "learning_rate": 2.2573176394944328e-05,
      "loss": 1.3815,
      "step": 2150
    },
    {
      "epoch": 0.24750773461670678,
      "grad_norm": 0.5789175629615784,
      "learning_rate": 2.2088605193112383e-05,
      "loss": 1.5056,
      "step": 2160
    },
    {
      "epoch": 0.24865360375845078,
      "grad_norm": 0.5871517658233643,
      "learning_rate": 2.160781274734495e-05,
      "loss": 1.6562,
      "step": 2170
    },
    {
      "epoch": 0.2497994729001948,
      "grad_norm": 0.5915065407752991,
      "learning_rate": 2.1130864149242878e-05,
      "loss": 1.6193,
      "step": 2180
    },
    {
      "epoch": 0.2509453420419388,
      "grad_norm": 0.6385588049888611,
      "learning_rate": 2.0657823970011618e-05,
      "loss": 1.5005,
      "step": 2190
    },
    {
      "epoch": 0.25209121118368283,
      "grad_norm": 0.6502992510795593,
      "learning_rate": 2.0188756251719203e-05,
      "loss": 1.5248,
      "step": 2200
    },
    {
      "epoch": 0.25323708032542686,
      "grad_norm": 0.5153409838676453,
      "learning_rate": 1.9723724498626105e-05,
      "loss": 1.5183,
      "step": 2210
    },
    {
      "epoch": 0.2543829494671708,
      "grad_norm": 0.5918806791305542,
      "learning_rate": 1.9262791668587676e-05,
      "loss": 1.3214,
      "step": 2220
    },
    {
      "epoch": 0.25552881860891485,
      "grad_norm": 0.7003433704376221,
      "learning_rate": 1.8806020164530702e-05,
      "loss": 1.6252,
      "step": 2230
    },
    {
      "epoch": 0.2566746877506589,
      "grad_norm": 0.5324897766113281,
      "learning_rate": 1.8353471826005036e-05,
      "loss": 1.3705,
      "step": 2240
    },
    {
      "epoch": 0.2578205568924029,
      "grad_norm": 0.49035802483558655,
      "learning_rate": 1.7905207920811572e-05,
      "loss": 1.4218,
      "step": 2250
    },
    {
      "epoch": 0.2589664260341469,
      "grad_norm": 0.651566207408905,
      "learning_rate": 1.746128913670746e-05,
      "loss": 1.3757,
      "step": 2260
    },
    {
      "epoch": 0.2601122951758909,
      "grad_norm": 0.6599149703979492,
      "learning_rate": 1.7021775573190013e-05,
      "loss": 1.4358,
      "step": 2270
    },
    {
      "epoch": 0.26125816431763493,
      "grad_norm": 0.7733114361763,
      "learning_rate": 1.6586726733360237e-05,
      "loss": 1.6378,
      "step": 2280
    },
    {
      "epoch": 0.26240403345937896,
      "grad_norm": 0.636096179485321,
      "learning_rate": 1.615620151586697e-05,
      "loss": 1.3983,
      "step": 2290
    },
    {
      "epoch": 0.26354990260112293,
      "grad_norm": 0.5301023125648499,
      "learning_rate": 1.5730258206933025e-05,
      "loss": 1.567,
      "step": 2300
    },
    {
      "epoch": 0.26469577174286696,
      "grad_norm": 0.5208801627159119,
      "learning_rate": 1.530895447246411e-05,
      "loss": 1.4182,
      "step": 2310
    },
    {
      "epoch": 0.265841640884611,
      "grad_norm": 0.6123260259628296,
      "learning_rate": 1.4892347350241881e-05,
      "loss": 1.6462,
      "step": 2320
    },
    {
      "epoch": 0.266987510026355,
      "grad_norm": 0.5237992405891418,
      "learning_rate": 1.448049324220181e-05,
      "loss": 1.4964,
      "step": 2330
    },
    {
      "epoch": 0.268133379168099,
      "grad_norm": 0.7907554507255554,
      "learning_rate": 1.4073447906797376e-05,
      "loss": 1.3589,
      "step": 2340
    },
    {
      "epoch": 0.269279248309843,
      "grad_norm": 0.6511983871459961,
      "learning_rate": 1.367126645145121e-05,
      "loss": 1.5064,
      "step": 2350
    },
    {
      "epoch": 0.27042511745158704,
      "grad_norm": 0.5229982137680054,
      "learning_rate": 1.327400332509442e-05,
      "loss": 1.3212,
      "step": 2360
    },
    {
      "epoch": 0.27157098659333107,
      "grad_norm": 0.6975625157356262,
      "learning_rate": 1.2881712310795118e-05,
      "loss": 1.1916,
      "step": 2370
    },
    {
      "epoch": 0.27271685573507504,
      "grad_norm": 0.7777551412582397,
      "learning_rate": 1.2494446518477022e-05,
      "loss": 1.476,
      "step": 2380
    },
    {
      "epoch": 0.27386272487681906,
      "grad_norm": 0.7115191221237183,
      "learning_rate": 1.2112258377729274e-05,
      "loss": 1.6018,
      "step": 2390
    },
    {
      "epoch": 0.2750085940185631,
      "grad_norm": 0.6283124685287476,
      "learning_rate": 1.1735199630708222e-05,
      "loss": 1.3557,
      "step": 2400
    },
    {
      "epoch": 0.2761544631603071,
      "grad_norm": 0.6109066605567932,
      "learning_rate": 1.1363321325132447e-05,
      "loss": 1.6158,
      "step": 2410
    },
    {
      "epoch": 0.2773003323020511,
      "grad_norm": 0.5065355896949768,
      "learning_rate": 1.0996673807371677e-05,
      "loss": 1.486,
      "step": 2420
    },
    {
      "epoch": 0.2784462014437951,
      "grad_norm": 0.5855929851531982,
      "learning_rate": 1.0635306715630682e-05,
      "loss": 1.5457,
      "step": 2430
    },
    {
      "epoch": 0.27959207058553914,
      "grad_norm": 0.5629914999008179,
      "learning_rate": 1.0279268973229089e-05,
      "loss": 1.6582,
      "step": 2440
    },
    {
      "epoch": 0.28073793972728317,
      "grad_norm": 0.5076742172241211,
      "learning_rate": 9.928608781977966e-06,
      "loss": 1.4035,
      "step": 2450
    },
    {
      "epoch": 0.28188380886902714,
      "grad_norm": 0.6038351058959961,
      "learning_rate": 9.583373615653978e-06,
      "loss": 1.4946,
      "step": 2460
    },
    {
      "epoch": 0.28302967801077117,
      "grad_norm": 0.6158405542373657,
      "learning_rate": 9.243610213572285e-06,
      "loss": 1.5954,
      "step": 2470
    },
    {
      "epoch": 0.2841755471525152,
      "grad_norm": 0.8014406561851501,
      "learning_rate": 8.909364574258793e-06,
      "loss": 1.6093,
      "step": 2480
    },
    {
      "epoch": 0.2853214162942592,
      "grad_norm": 0.6260013580322266,
      "learning_rate": 8.580681949222568e-06,
      "loss": 1.5621,
      "step": 2490
    },
    {
      "epoch": 0.2864672854360032,
      "grad_norm": 0.5818029046058655,
      "learning_rate": 8.257606836829678e-06,
      "loss": 1.4025,
      "step": 2500
    },
    {
      "epoch": 0.2876131545777472,
      "grad_norm": 0.6822273135185242,
      "learning_rate": 7.940182976278692e-06,
      "loss": 1.3993,
      "step": 2510
    },
    {
      "epoch": 0.28875902371949125,
      "grad_norm": 0.5265905261039734,
      "learning_rate": 7.6284533416791814e-06,
      "loss": 1.5283,
      "step": 2520
    },
    {
      "epoch": 0.2899048928612352,
      "grad_norm": 0.6255074739456177,
      "learning_rate": 7.322460136233622e-06,
      "loss": 1.5452,
      "step": 2530
    },
    {
      "epoch": 0.29105076200297925,
      "grad_norm": 0.5332224369049072,
      "learning_rate": 7.02224478652388e-06,
      "loss": 1.4633,
      "step": 2540
    },
    {
      "epoch": 0.2921966311447233,
      "grad_norm": 0.6171149015426636,
      "learning_rate": 6.727847936902543e-06,
      "loss": 1.4969,
      "step": 2550
    },
    {
      "epoch": 0.2933425002864673,
      "grad_norm": 0.5771318674087524,
      "learning_rate": 6.439309443990532e-06,
      "loss": 1.613,
      "step": 2560
    },
    {
      "epoch": 0.29448836942821127,
      "grad_norm": 0.4912353754043579,
      "learning_rate": 6.1566683712809824e-06,
      "loss": 1.3952,
      "step": 2570
    },
    {
      "epoch": 0.2956342385699553,
      "grad_norm": 0.6518236994743347,
      "learning_rate": 5.879962983850745e-06,
      "loss": 1.4907,
      "step": 2580
    },
    {
      "epoch": 0.2967801077116993,
      "grad_norm": 0.47168079018592834,
      "learning_rate": 5.6092307431799384e-06,
      "loss": 1.7893,
      "step": 2590
    },
    {
      "epoch": 0.29792597685344335,
      "grad_norm": 0.6220501661300659,
      "learning_rate": 5.34450830208017e-06,
      "loss": 1.2903,
      "step": 2600
    },
    {
      "epoch": 0.2990718459951873,
      "grad_norm": 0.6064465641975403,
      "learning_rate": 5.08583149973243e-06,
      "loss": 1.4964,
      "step": 2610
    },
    {
      "epoch": 0.30021771513693135,
      "grad_norm": 0.5346307158470154,
      "learning_rate": 4.833235356834959e-06,
      "loss": 1.4922,
      "step": 2620
    },
    {
      "epoch": 0.3013635842786754,
      "grad_norm": 0.6800510883331299,
      "learning_rate": 4.586754070862099e-06,
      "loss": 1.4117,
      "step": 2630
    },
    {
      "epoch": 0.3025094534204194,
      "grad_norm": 0.5367140769958496,
      "learning_rate": 4.346421011434382e-06,
      "loss": 1.606,
      "step": 2640
    },
    {
      "epoch": 0.3036553225621634,
      "grad_norm": 0.5802227854728699,
      "learning_rate": 4.112268715800943e-06,
      "loss": 1.3678,
      "step": 2650
    },
    {
      "epoch": 0.3048011917039074,
      "grad_norm": 0.7918272614479065,
      "learning_rate": 3.884328884434402e-06,
      "loss": 1.4701,
      "step": 2660
    },
    {
      "epoch": 0.30594706084565143,
      "grad_norm": 0.5996667742729187,
      "learning_rate": 3.6626323767391777e-06,
      "loss": 1.4663,
      "step": 2670
    },
    {
      "epoch": 0.30709292998739546,
      "grad_norm": 0.586614727973938,
      "learning_rate": 3.4472092068735916e-06,
      "loss": 1.4146,
      "step": 2680
    },
    {
      "epoch": 0.30823879912913943,
      "grad_norm": 0.6410096287727356,
      "learning_rate": 3.238088539686451e-06,
      "loss": 1.47,
      "step": 2690
    },
    {
      "epoch": 0.30938466827088346,
      "grad_norm": 0.5319120287895203,
      "learning_rate": 3.0352986867686007e-06,
      "loss": 1.6689,
      "step": 2700
    },
    {
      "epoch": 0.3105305374126275,
      "grad_norm": 0.641806960105896,
      "learning_rate": 2.8388671026199522e-06,
      "loss": 1.5412,
      "step": 2710
    },
    {
      "epoch": 0.3116764065543715,
      "grad_norm": 0.6084569096565247,
      "learning_rate": 2.6488203809326207e-06,
      "loss": 1.3932,
      "step": 2720
    },
    {
      "epoch": 0.3128222756961155,
      "grad_norm": 0.5853345990180969,
      "learning_rate": 2.4651842509905487e-06,
      "loss": 1.4632,
      "step": 2730
    },
    {
      "epoch": 0.3139681448378595,
      "grad_norm": 0.5775859951972961,
      "learning_rate": 2.2879835741861586e-06,
      "loss": 1.3812,
      "step": 2740
    },
    {
      "epoch": 0.31511401397960354,
      "grad_norm": 0.5420204997062683,
      "learning_rate": 2.1172423406545516e-06,
      "loss": 1.5721,
      "step": 2750
    },
    {
      "epoch": 0.31625988312134756,
      "grad_norm": 0.48948049545288086,
      "learning_rate": 1.9529836660256096e-06,
      "loss": 1.6311,
      "step": 2760
    },
    {
      "epoch": 0.31740575226309153,
      "grad_norm": 0.6526772975921631,
      "learning_rate": 1.7952297882945003e-06,
      "loss": 1.7197,
      "step": 2770
    },
    {
      "epoch": 0.31855162140483556,
      "grad_norm": 0.5312083959579468,
      "learning_rate": 1.6440020648110067e-06,
      "loss": 1.3684,
      "step": 2780
    },
    {
      "epoch": 0.3196974905465796,
      "grad_norm": 0.6139872074127197,
      "learning_rate": 1.4993209693881183e-06,
      "loss": 1.5396,
      "step": 2790
    },
    {
      "epoch": 0.3208433596883236,
      "grad_norm": 0.6308534741401672,
      "learning_rate": 1.3612060895301759e-06,
      "loss": 1.4659,
      "step": 2800
    },
    {
      "epoch": 0.3219892288300676,
      "grad_norm": 0.5623085498809814,
      "learning_rate": 1.2296761237810207e-06,
      "loss": 1.4104,
      "step": 2810
    },
    {
      "epoch": 0.3231350979718116,
      "grad_norm": 0.5706859827041626,
      "learning_rate": 1.104748879192552e-06,
      "loss": 1.5481,
      "step": 2820
    },
    {
      "epoch": 0.32428096711355564,
      "grad_norm": 0.5466167330741882,
      "learning_rate": 9.864412689139123e-07,
      "loss": 1.4769,
      "step": 2830
    },
    {
      "epoch": 0.32542683625529967,
      "grad_norm": 0.5880963802337646,
      "learning_rate": 8.747693099017129e-07,
      "loss": 1.5516,
      "step": 2840
    },
    {
      "epoch": 0.32657270539704364,
      "grad_norm": 0.664244532585144,
      "learning_rate": 7.697481207516289e-07,
      "loss": 1.2943,
      "step": 2850
    },
    {
      "epoch": 0.32771857453878767,
      "grad_norm": 0.8473872542381287,
      "learning_rate": 6.713919196515317e-07,
      "loss": 1.2507,
      "step": 2860
    },
    {
      "epoch": 0.3288644436805317,
      "grad_norm": 0.8418035507202148,
      "learning_rate": 5.797140224566122e-07,
      "loss": 1.4377,
      "step": 2870
    },
    {
      "epoch": 0.3300103128222757,
      "grad_norm": 0.6410139203071594,
      "learning_rate": 4.947268408866113e-07,
      "loss": 1.4249,
      "step": 2880
    },
    {
      "epoch": 0.3311561819640197,
      "grad_norm": 0.6374297142028809,
      "learning_rate": 4.1644188084548063e-07,
      "loss": 1.5552,
      "step": 2890
    },
    {
      "epoch": 0.3323020511057637,
      "grad_norm": 0.701400637626648,
      "learning_rate": 3.4486974086366253e-07,
      "loss": 1.5406,
      "step": 2900
    },
    {
      "epoch": 0.33344792024750775,
      "grad_norm": 0.6102722883224487,
      "learning_rate": 2.800201106632205e-07,
      "loss": 1.5778,
      "step": 2910
    },
    {
      "epoch": 0.3345937893892518,
      "grad_norm": 0.6845527291297913,
      "learning_rate": 2.219017698460002e-07,
      "loss": 1.5503,
      "step": 2920
    },
    {
      "epoch": 0.33573965853099574,
      "grad_norm": 0.5155937075614929,
      "learning_rate": 1.7052258670501308e-07,
      "loss": 1.4669,
      "step": 2930
    },
    {
      "epoch": 0.33688552767273977,
      "grad_norm": 0.8121209740638733,
      "learning_rate": 1.2588951715921116e-07,
      "loss": 1.4042,
      "step": 2940
    },
    {
      "epoch": 0.3380313968144838,
      "grad_norm": 0.8292280435562134,
      "learning_rate": 8.800860381173448e-08,
      "loss": 1.415,
      "step": 2950
    },
    {
      "epoch": 0.3391772659562278,
      "grad_norm": 0.6214461326599121,
      "learning_rate": 5.688497513188229e-08,
      "loss": 1.2916,
      "step": 2960
    },
    {
      "epoch": 0.3403231350979718,
      "grad_norm": 0.6468619704246521,
      "learning_rate": 3.2522844760762836e-08,
      "loss": 1.3698,
      "step": 2970
    },
    {
      "epoch": 0.3414690042397158,
      "grad_norm": 0.5640342831611633,
      "learning_rate": 1.4925510940844156e-08,
      "loss": 1.336,
      "step": 2980
    },
    {
      "epoch": 0.34261487338145985,
      "grad_norm": 0.4522778391838074,
      "learning_rate": 4.095356069439005e-09,
      "loss": 1.3447,
      "step": 2990
    },
    {
      "epoch": 0.3437607425232039,
      "grad_norm": 0.560656726360321,
      "learning_rate": 3.384637615733155e-11,
      "loss": 1.4884,
      "step": 3000
    }
  ],
  "logging_steps": 10,
  "max_steps": 3000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 300,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.91880651046912e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}