{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 2111,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.014214641080312722,
      "grad_norm": 2.834075681079323,
      "learning_rate": 4.2452830188679244e-07,
      "loss": 0.8888,
      "step": 10
    },
    {
      "epoch": 0.028429282160625444,
      "grad_norm": 2.382206917176712,
      "learning_rate": 8.962264150943397e-07,
      "loss": 0.8863,
      "step": 20
    },
    {
      "epoch": 0.042643923240938165,
      "grad_norm": 1.5553139981482849,
      "learning_rate": 1.3679245283018869e-06,
      "loss": 0.8307,
      "step": 30
    },
    {
      "epoch": 0.05685856432125089,
      "grad_norm": 1.2537358335844357,
      "learning_rate": 1.839622641509434e-06,
      "loss": 0.7826,
      "step": 40
    },
    {
      "epoch": 0.07107320540156362,
      "grad_norm": 0.766804883609826,
      "learning_rate": 2.3113207547169815e-06,
      "loss": 0.7347,
      "step": 50
    },
    {
      "epoch": 0.08528784648187633,
      "grad_norm": 0.7667121953750998,
      "learning_rate": 2.7830188679245286e-06,
      "loss": 0.6931,
      "step": 60
    },
    {
      "epoch": 0.09950248756218906,
      "grad_norm": 0.6131128302034111,
      "learning_rate": 3.2547169811320758e-06,
      "loss": 0.6639,
      "step": 70
    },
    {
      "epoch": 0.11371712864250177,
      "grad_norm": 0.6988128095327644,
      "learning_rate": 3.726415094339623e-06,
      "loss": 0.6375,
      "step": 80
    },
    {
      "epoch": 0.1279317697228145,
      "grad_norm": 0.6317415794697651,
      "learning_rate": 4.19811320754717e-06,
      "loss": 0.6236,
      "step": 90
    },
    {
      "epoch": 0.14214641080312723,
      "grad_norm": 0.6086478743343868,
      "learning_rate": 4.6698113207547175e-06,
      "loss": 0.6138,
      "step": 100
    },
    {
      "epoch": 0.15636105188343993,
      "grad_norm": 0.5886423623895793,
      "learning_rate": 5.1415094339622655e-06,
      "loss": 0.6086,
      "step": 110
    },
    {
      "epoch": 0.17057569296375266,
      "grad_norm": 0.6688530721380745,
      "learning_rate": 5.613207547169813e-06,
      "loss": 0.603,
      "step": 120
    },
    {
      "epoch": 0.1847903340440654,
      "grad_norm": 0.6884356457949358,
      "learning_rate": 6.08490566037736e-06,
      "loss": 0.5903,
      "step": 130
    },
    {
      "epoch": 0.19900497512437812,
      "grad_norm": 0.609548931554606,
      "learning_rate": 6.556603773584907e-06,
      "loss": 0.5966,
      "step": 140
    },
    {
      "epoch": 0.21321961620469082,
      "grad_norm": 0.5949327566115654,
      "learning_rate": 7.028301886792454e-06,
      "loss": 0.5904,
      "step": 150
    },
    {
      "epoch": 0.22743425728500355,
      "grad_norm": 0.6242334967031955,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.579,
      "step": 160
    },
    {
      "epoch": 0.24164889836531628,
      "grad_norm": 0.6241786288297929,
      "learning_rate": 7.971698113207547e-06,
      "loss": 0.5791,
      "step": 170
    },
    {
      "epoch": 0.255863539445629,
      "grad_norm": 0.6521923298448227,
      "learning_rate": 8.443396226415095e-06,
      "loss": 0.5772,
      "step": 180
    },
    {
      "epoch": 0.27007818052594174,
      "grad_norm": 0.72406379743765,
      "learning_rate": 8.915094339622642e-06,
      "loss": 0.5783,
      "step": 190
    },
    {
      "epoch": 0.28429282160625446,
      "grad_norm": 0.6865581822637807,
      "learning_rate": 9.38679245283019e-06,
      "loss": 0.5667,
      "step": 200
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.7370553596384785,
      "learning_rate": 9.858490566037736e-06,
      "loss": 0.5787,
      "step": 210
    },
    {
      "epoch": 0.31272210376687987,
      "grad_norm": 0.680822415396935,
      "learning_rate": 9.999665093340166e-06,
      "loss": 0.5693,
      "step": 220
    },
    {
      "epoch": 0.3269367448471926,
      "grad_norm": 0.7111908345205171,
      "learning_rate": 9.998024842193876e-06,
      "loss": 0.5513,
      "step": 230
    },
    {
      "epoch": 0.3411513859275053,
      "grad_norm": 0.7236426819082362,
      "learning_rate": 9.995018180960701e-06,
      "loss": 0.5713,
      "step": 240
    },
    {
      "epoch": 0.35536602700781805,
      "grad_norm": 0.5852089481414202,
      "learning_rate": 9.990645931631796e-06,
      "loss": 0.5538,
      "step": 250
    },
    {
      "epoch": 0.3695806680881308,
      "grad_norm": 0.6824815552672151,
      "learning_rate": 9.984909289536473e-06,
      "loss": 0.5532,
      "step": 260
    },
    {
      "epoch": 0.3837953091684435,
      "grad_norm": 0.6415502778484192,
      "learning_rate": 9.9778098230154e-06,
      "loss": 0.5651,
      "step": 270
    },
    {
      "epoch": 0.39800995024875624,
      "grad_norm": 0.6951097230627817,
      "learning_rate": 9.969349472991838e-06,
      "loss": 0.5545,
      "step": 280
    },
    {
      "epoch": 0.41222459132906897,
      "grad_norm": 0.6940350949391038,
      "learning_rate": 9.959530552441006e-06,
      "loss": 0.5618,
      "step": 290
    },
    {
      "epoch": 0.42643923240938164,
      "grad_norm": 0.6795280607455072,
      "learning_rate": 9.94835574575774e-06,
      "loss": 0.5678,
      "step": 300
    },
    {
      "epoch": 0.44065387348969437,
      "grad_norm": 0.740615761131801,
      "learning_rate": 9.93582810802261e-06,
      "loss": 0.5589,
      "step": 310
    },
    {
      "epoch": 0.4548685145700071,
      "grad_norm": 0.6190142422707086,
      "learning_rate": 9.921951064166685e-06,
      "loss": 0.5545,
      "step": 320
    },
    {
      "epoch": 0.4690831556503198,
      "grad_norm": 0.6133098155280035,
      "learning_rate": 9.90672840803519e-06,
      "loss": 0.5424,
      "step": 330
    },
    {
      "epoch": 0.48329779673063256,
      "grad_norm": 0.6671428298287642,
      "learning_rate": 9.890164301350318e-06,
      "loss": 0.5543,
      "step": 340
    },
    {
      "epoch": 0.4975124378109453,
      "grad_norm": 0.6195386955465261,
      "learning_rate": 9.872263272573443e-06,
      "loss": 0.5436,
      "step": 350
    },
    {
      "epoch": 0.511727078891258,
      "grad_norm": 0.5930299278047146,
      "learning_rate": 9.853030215667095e-06,
      "loss": 0.5419,
      "step": 360
    },
    {
      "epoch": 0.5259417199715707,
      "grad_norm": 0.5808813232166387,
      "learning_rate": 9.832470388756987e-06,
      "loss": 0.5355,
      "step": 370
    },
    {
      "epoch": 0.5401563610518835,
      "grad_norm": 0.6451836622188146,
      "learning_rate": 9.81058941269451e-06,
      "loss": 0.5411,
      "step": 380
    },
    {
      "epoch": 0.5543710021321961,
      "grad_norm": 0.6430253547664657,
      "learning_rate": 9.787393269520039e-06,
      "loss": 0.5424,
      "step": 390
    },
    {
      "epoch": 0.5685856432125089,
      "grad_norm": 0.6434468036531783,
      "learning_rate": 9.762888300827507e-06,
      "loss": 0.5375,
      "step": 400
    },
    {
      "epoch": 0.5828002842928216,
      "grad_norm": 0.6545071562387185,
      "learning_rate": 9.737081206030671e-06,
      "loss": 0.5392,
      "step": 410
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.6520857107905386,
      "learning_rate": 9.709979040531568e-06,
      "loss": 0.5394,
      "step": 420
    },
    {
      "epoch": 0.6112295664534471,
      "grad_norm": 0.6850212115229948,
      "learning_rate": 9.681589213791633e-06,
      "loss": 0.5466,
      "step": 430
    },
    {
      "epoch": 0.6254442075337597,
      "grad_norm": 0.6269054102432758,
      "learning_rate": 9.651919487306025e-06,
      "loss": 0.5322,
      "step": 440
    },
    {
      "epoch": 0.6396588486140725,
      "grad_norm": 0.736787330144809,
      "learning_rate": 9.620977972481715e-06,
      "loss": 0.5429,
      "step": 450
    },
    {
      "epoch": 0.6538734896943852,
      "grad_norm": 0.662925500196175,
      "learning_rate": 9.588773128419907e-06,
      "loss": 0.5312,
      "step": 460
    },
    {
      "epoch": 0.668088130774698,
      "grad_norm": 0.6148143038929397,
      "learning_rate": 9.555313759603403e-06,
      "loss": 0.5316,
      "step": 470
    },
    {
      "epoch": 0.6823027718550106,
      "grad_norm": 0.6077969318281708,
      "learning_rate": 9.520609013489548e-06,
      "loss": 0.5223,
      "step": 480
    },
    {
      "epoch": 0.6965174129353234,
      "grad_norm": 0.6168040167150479,
      "learning_rate": 9.484668378009407e-06,
      "loss": 0.5405,
      "step": 490
    },
    {
      "epoch": 0.7107320540156361,
      "grad_norm": 0.6128096552430267,
      "learning_rate": 9.447501678973853e-06,
      "loss": 0.5338,
      "step": 500
    },
    {
      "epoch": 0.7107320540156361,
      "eval_loss": 0.5341117978096008,
      "eval_runtime": 234.2375,
      "eval_samples_per_second": 21.346,
      "eval_steps_per_second": 2.668,
      "step": 500
    },
    {
      "epoch": 0.7249466950959488,
      "grad_norm": 0.6234640184402841,
      "learning_rate": 9.409119077387295e-06,
      "loss": 0.5377,
      "step": 510
    },
    {
      "epoch": 0.7391613361762616,
      "grad_norm": 0.6131203844453222,
      "learning_rate": 9.369531066669759e-06,
      "loss": 0.5259,
      "step": 520
    },
    {
      "epoch": 0.7533759772565742,
      "grad_norm": 0.6262740746023064,
      "learning_rate": 9.328748469788094e-06,
      "loss": 0.5328,
      "step": 530
    },
    {
      "epoch": 0.767590618336887,
      "grad_norm": 0.6262806652099724,
      "learning_rate": 9.286782436297072e-06,
      "loss": 0.5344,
      "step": 540
    },
    {
      "epoch": 0.7818052594171997,
      "grad_norm": 0.6615647825526851,
      "learning_rate": 9.243644439291223e-06,
      "loss": 0.5257,
      "step": 550
    },
    {
      "epoch": 0.7960199004975125,
      "grad_norm": 0.6072216321225935,
      "learning_rate": 9.1993462722682e-06,
      "loss": 0.5339,
      "step": 560
    },
    {
      "epoch": 0.8102345415778252,
      "grad_norm": 0.5963884112129639,
      "learning_rate": 9.15390004590455e-06,
      "loss": 0.5235,
      "step": 570
    },
    {
      "epoch": 0.8244491826581379,
      "grad_norm": 0.6479032652506073,
      "learning_rate": 9.107318184744782e-06,
      "loss": 0.5244,
      "step": 580
    },
    {
      "epoch": 0.8386638237384506,
      "grad_norm": 0.6241228532450154,
      "learning_rate": 9.059613423804623e-06,
      "loss": 0.5279,
      "step": 590
    },
    {
      "epoch": 0.8528784648187633,
      "grad_norm": 0.6346658666864172,
      "learning_rate": 9.010798805089385e-06,
      "loss": 0.5315,
      "step": 600
    },
    {
      "epoch": 0.8670931058990761,
      "grad_norm": 0.6189100879810271,
      "learning_rate": 8.960887674028411e-06,
      "loss": 0.516,
      "step": 610
    },
    {
      "epoch": 0.8813077469793887,
      "grad_norm": 0.6310100666801625,
      "learning_rate": 8.909893675826575e-06,
      "loss": 0.5343,
      "step": 620
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.5705654281594257,
      "learning_rate": 8.857830751733815e-06,
      "loss": 0.5196,
      "step": 630
    },
    {
      "epoch": 0.9097370291400142,
      "grad_norm": 0.581798609611747,
      "learning_rate": 8.80471313523373e-06,
      "loss": 0.5252,
      "step": 640
    },
    {
      "epoch": 0.923951670220327,
      "grad_norm": 0.6021206642948125,
      "learning_rate": 8.750555348152299e-06,
      "loss": 0.5301,
      "step": 650
    },
    {
      "epoch": 0.9381663113006397,
      "grad_norm": 0.6258455776720466,
      "learning_rate": 8.695372196687743e-06,
      "loss": 0.5247,
      "step": 660
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 0.5984715139941876,
      "learning_rate": 8.639178767362677e-06,
      "loss": 0.5166,
      "step": 670
    },
    {
      "epoch": 0.9665955934612651,
      "grad_norm": 0.6174130149894701,
      "learning_rate": 8.581990422899586e-06,
      "loss": 0.5186,
      "step": 680
    },
    {
      "epoch": 0.9808102345415778,
      "grad_norm": 0.669919825928536,
      "learning_rate": 8.523822798020827e-06,
      "loss": 0.5232,
      "step": 690
    },
    {
      "epoch": 0.9950248756218906,
      "grad_norm": 0.5789470998307686,
      "learning_rate": 8.46469179517424e-06,
      "loss": 0.5128,
      "step": 700
    },
    {
      "epoch": 1.0085287846481876,
      "grad_norm": 0.6391934862223132,
      "learning_rate": 8.404613580185586e-06,
      "loss": 0.4571,
      "step": 710
    },
    {
      "epoch": 1.0227434257285004,
      "grad_norm": 0.646088830119043,
      "learning_rate": 8.343604577838965e-06,
      "loss": 0.4844,
      "step": 720
    },
    {
      "epoch": 1.036958066808813,
      "grad_norm": 0.5785525652959986,
      "learning_rate": 8.281681467386447e-06,
      "loss": 0.4731,
      "step": 730
    },
    {
      "epoch": 1.0511727078891258,
      "grad_norm": 0.6609376724143528,
      "learning_rate": 8.21886117798813e-06,
      "loss": 0.483,
      "step": 740
    },
    {
      "epoch": 1.0653873489694385,
      "grad_norm": 0.6092591984384597,
      "learning_rate": 8.155160884083881e-06,
      "loss": 0.4868,
      "step": 750
    },
    {
      "epoch": 1.0796019900497513,
      "grad_norm": 0.6374135491483588,
      "learning_rate": 8.090598000698009e-06,
      "loss": 0.492,
      "step": 760
    },
    {
      "epoch": 1.0938166311300639,
      "grad_norm": 0.6008076415379893,
      "learning_rate": 8.025190178678175e-06,
      "loss": 0.4814,
      "step": 770
    },
    {
      "epoch": 1.1080312722103767,
      "grad_norm": 0.6018194277584734,
      "learning_rate": 7.958955299869826e-06,
      "loss": 0.4768,
      "step": 780
    },
    {
      "epoch": 1.1222459132906895,
      "grad_norm": 0.5917444038650317,
      "learning_rate": 7.891911472227478e-06,
      "loss": 0.4841,
      "step": 790
    },
    {
      "epoch": 1.136460554371002,
      "grad_norm": 0.6728799540384178,
      "learning_rate": 7.82407702486418e-06,
      "loss": 0.4797,
      "step": 800
    },
    {
      "epoch": 1.1506751954513148,
      "grad_norm": 0.5791378894590185,
      "learning_rate": 7.755470503040516e-06,
      "loss": 0.4842,
      "step": 810
    },
    {
      "epoch": 1.1648898365316276,
      "grad_norm": 0.6318987068220235,
      "learning_rate": 7.686110663094527e-06,
      "loss": 0.4752,
      "step": 820
    },
    {
      "epoch": 1.1791044776119404,
      "grad_norm": 0.5836896353182887,
      "learning_rate": 7.616016467313891e-06,
      "loss": 0.4783,
      "step": 830
    },
    {
      "epoch": 1.193319118692253,
      "grad_norm": 0.6133166239448045,
      "learning_rate": 7.545207078751858e-06,
      "loss": 0.4804,
      "step": 840
    },
    {
      "epoch": 1.2075337597725657,
      "grad_norm": 0.6647464026501058,
      "learning_rate": 7.473701855988227e-06,
      "loss": 0.4837,
      "step": 850
    },
    {
      "epoch": 1.2217484008528785,
      "grad_norm": 0.6037242762628189,
      "learning_rate": 7.4015203478369266e-06,
      "loss": 0.48,
      "step": 860
    },
    {
      "epoch": 1.235963041933191,
      "grad_norm": 0.6060908786113385,
      "learning_rate": 7.328682288001561e-06,
      "loss": 0.4828,
      "step": 870
    },
    {
      "epoch": 1.2501776830135038,
      "grad_norm": 0.5748244127007777,
      "learning_rate": 7.255207589680403e-06,
      "loss": 0.4737,
      "step": 880
    },
    {
      "epoch": 1.2643923240938166,
      "grad_norm": 0.5739667849410425,
      "learning_rate": 7.181116340122336e-06,
      "loss": 0.4692,
      "step": 890
    },
    {
      "epoch": 1.2786069651741294,
      "grad_norm": 0.5356752158374642,
      "learning_rate": 7.10642879513519e-06,
      "loss": 0.467,
      "step": 900
    },
    {
      "epoch": 1.2928216062544422,
      "grad_norm": 0.6231787580072905,
      "learning_rate": 7.0311653735480136e-06,
      "loss": 0.4787,
      "step": 910
    },
    {
      "epoch": 1.3070362473347548,
      "grad_norm": 0.5539031871696426,
      "learning_rate": 6.95534665162877e-06,
      "loss": 0.4803,
      "step": 920
    },
    {
      "epoch": 1.3212508884150675,
      "grad_norm": 0.5545024040236839,
      "learning_rate": 6.878993357458986e-06,
      "loss": 0.4727,
      "step": 930
    },
    {
      "epoch": 1.33546552949538,
      "grad_norm": 0.5526343563080722,
      "learning_rate": 6.8021263652669055e-06,
      "loss": 0.4749,
      "step": 940
    },
    {
      "epoch": 1.349680170575693,
      "grad_norm": 0.5949065425774154,
      "learning_rate": 6.7247666897206795e-06,
      "loss": 0.466,
      "step": 950
    },
    {
      "epoch": 1.3638948116560057,
      "grad_norm": 0.6109499791395212,
      "learning_rate": 6.646935480183173e-06,
      "loss": 0.4721,
      "step": 960
    },
    {
      "epoch": 1.3781094527363185,
      "grad_norm": 0.5761776478178481,
      "learning_rate": 6.568654014929933e-06,
      "loss": 0.4773,
      "step": 970
    },
    {
      "epoch": 1.3923240938166312,
      "grad_norm": 0.6309785557491375,
      "learning_rate": 6.4899436953319235e-06,
      "loss": 0.4783,
      "step": 980
    },
    {
      "epoch": 1.4065387348969438,
      "grad_norm": 0.6079839523261087,
      "learning_rate": 6.410826040004607e-06,
      "loss": 0.4874,
      "step": 990
    },
    {
      "epoch": 1.4207533759772566,
      "grad_norm": 0.5694417979836612,
      "learning_rate": 6.331322678924963e-06,
      "loss": 0.4883,
      "step": 1000
    },
    {
      "epoch": 1.4207533759772566,
      "eval_loss": 0.5142297148704529,
      "eval_runtime": 233.9812,
      "eval_samples_per_second": 21.369,
      "eval_steps_per_second": 2.671,
      "step": 1000
    },
    {
      "epoch": 1.4349680170575694,
      "grad_norm": 0.590623858748834,
      "learning_rate": 6.251455347518074e-06,
      "loss": 0.4674,
      "step": 1010
    },
    {
      "epoch": 1.449182658137882,
      "grad_norm": 0.6492265919143085,
      "learning_rate": 6.1712458807148804e-06,
      "loss": 0.4566,
      "step": 1020
    },
    {
      "epoch": 1.4633972992181947,
      "grad_norm": 0.6690825078821129,
      "learning_rate": 6.090716206982714e-06,
      "loss": 0.4415,
      "step": 1030
    },
    {
      "epoch": 1.4776119402985075,
      "grad_norm": 0.6207866027112856,
      "learning_rate": 6.009888342330292e-06,
      "loss": 0.4419,
      "step": 1040
    },
    {
      "epoch": 1.4918265813788203,
      "grad_norm": 0.5983146438640481,
      "learning_rate": 5.92878438428875e-06,
      "loss": 0.4529,
      "step": 1050
    },
    {
      "epoch": 1.5060412224591329,
      "grad_norm": 0.6251595437559933,
      "learning_rate": 5.847426505870399e-06,
      "loss": 0.4433,
      "step": 1060
    },
    {
      "epoch": 1.5202558635394456,
      "grad_norm": 0.5973997328290039,
      "learning_rate": 5.765836949506843e-06,
      "loss": 0.4465,
      "step": 1070
    },
    {
      "epoch": 1.5344705046197582,
      "grad_norm": 0.61605264507367,
      "learning_rate": 5.684038020968126e-06,
      "loss": 0.4354,
      "step": 1080
    },
    {
      "epoch": 1.548685145700071,
      "grad_norm": 0.6356804350566382,
      "learning_rate": 5.6020520832645555e-06,
      "loss": 0.4474,
      "step": 1090
    },
    {
      "epoch": 1.5628997867803838,
      "grad_norm": 0.5409307423191971,
      "learning_rate": 5.519901550532871e-06,
      "loss": 0.4442,
      "step": 1100
    },
    {
      "epoch": 1.5771144278606966,
      "grad_norm": 0.5628516264470729,
      "learning_rate": 5.437608881908456e-06,
      "loss": 0.4469,
      "step": 1110
    },
    {
      "epoch": 1.5913290689410093,
      "grad_norm": 0.5646776831646609,
      "learning_rate": 5.3551965753852255e-06,
      "loss": 0.4444,
      "step": 1120
    },
    {
      "epoch": 1.6055437100213221,
      "grad_norm": 0.6070095692660494,
      "learning_rate": 5.2726871616649e-06,
      "loss": 0.4546,
      "step": 1130
    },
    {
      "epoch": 1.6197583511016347,
      "grad_norm": 0.5955403854226203,
      "learning_rate": 5.190103197997339e-06,
      "loss": 0.4453,
      "step": 1140
    },
    {
      "epoch": 1.6339729921819472,
      "grad_norm": 0.5847049727426836,
      "learning_rate": 5.107467262013614e-06,
      "loss": 0.4456,
      "step": 1150
    },
    {
      "epoch": 1.64818763326226,
      "grad_norm": 0.6464335190947482,
      "learning_rate": 5.02480194555351e-06,
      "loss": 0.4431,
      "step": 1160
    },
    {
      "epoch": 1.6624022743425728,
      "grad_norm": 0.5976768465351757,
      "learning_rate": 4.942129848489137e-06,
      "loss": 0.4459,
      "step": 1170
    },
    {
      "epoch": 1.6766169154228856,
      "grad_norm": 0.5759461226301025,
      "learning_rate": 4.8594735725463575e-06,
      "loss": 0.4331,
      "step": 1180
    },
    {
      "epoch": 1.6908315565031984,
      "grad_norm": 0.5618161675831925,
      "learning_rate": 4.776855715125694e-06,
      "loss": 0.4459,
      "step": 1190
    },
    {
      "epoch": 1.7050461975835112,
      "grad_norm": 0.5665638890009612,
      "learning_rate": 4.694298863124435e-06,
      "loss": 0.4434,
      "step": 1200
    },
    {
      "epoch": 1.7192608386638237,
      "grad_norm": 0.5897127590620418,
      "learning_rate": 4.611825586761591e-06,
      "loss": 0.4526,
      "step": 1210
    },
    {
      "epoch": 1.7334754797441365,
      "grad_norm": 0.5964155647113037,
      "learning_rate": 4.529458433407429e-06,
      "loss": 0.4404,
      "step": 1220
    },
    {
      "epoch": 1.747690120824449,
      "grad_norm": 0.557832612527273,
      "learning_rate": 4.447219921419244e-06,
      "loss": 0.4466,
      "step": 1230
    },
    {
      "epoch": 1.7619047619047619,
      "grad_norm": 0.6542060047375909,
      "learning_rate": 4.365132533985071e-06,
      "loss": 0.4456,
      "step": 1240
    },
    {
      "epoch": 1.7761194029850746,
      "grad_norm": 0.5917352620945386,
      "learning_rate": 4.283218712976992e-06,
      "loss": 0.4452,
      "step": 1250
    },
    {
      "epoch": 1.7903340440653874,
      "grad_norm": 0.561379370088737,
      "learning_rate": 4.201500852815769e-06,
      "loss": 0.442,
      "step": 1260
    },
    {
      "epoch": 1.8045486851457002,
      "grad_norm": 0.5491467540898044,
      "learning_rate": 4.12000129434842e-06,
      "loss": 0.4444,
      "step": 1270
    },
    {
      "epoch": 1.8187633262260128,
      "grad_norm": 0.5872206166430523,
      "learning_rate": 4.0387423187404656e-06,
      "loss": 0.4375,
      "step": 1280
    },
    {
      "epoch": 1.8329779673063256,
      "grad_norm": 0.5927050301786448,
      "learning_rate": 3.957746141384469e-06,
      "loss": 0.4471,
      "step": 1290
    },
    {
      "epoch": 1.8471926083866381,
      "grad_norm": 0.5414882779351676,
      "learning_rate": 3.877034905826577e-06,
      "loss": 0.4466,
      "step": 1300
    },
    {
      "epoch": 1.861407249466951,
      "grad_norm": 0.5612180051513147,
      "learning_rate": 3.796630677712697e-06,
      "loss": 0.4388,
      "step": 1310
    },
    {
      "epoch": 1.8756218905472637,
      "grad_norm": 0.5529411457142649,
      "learning_rate": 3.716555438755961e-06,
      "loss": 0.4472,
      "step": 1320
    },
    {
      "epoch": 1.8898365316275765,
      "grad_norm": 0.5616746991079874,
      "learning_rate": 3.6368310807271546e-06,
      "loss": 0.4446,
      "step": 1330
    },
    {
      "epoch": 1.9040511727078893,
      "grad_norm": 0.5799752713225496,
      "learning_rate": 3.557479399469721e-06,
      "loss": 0.4493,
      "step": 1340
    },
    {
      "epoch": 1.9182658137882018,
      "grad_norm": 0.6619814912951459,
      "learning_rate": 3.4785220889409934e-06,
      "loss": 0.4405,
      "step": 1350
    },
    {
      "epoch": 1.9324804548685146,
      "grad_norm": 0.5613562889297112,
      "learning_rate": 3.3999807352812862e-06,
      "loss": 0.4541,
      "step": 1360
    },
    {
      "epoch": 1.9466950959488272,
      "grad_norm": 0.5713920903616314,
      "learning_rate": 3.321876810912461e-06,
      "loss": 0.4428,
      "step": 1370
    },
    {
      "epoch": 1.96090973702914,
      "grad_norm": 0.5665905645368096,
      "learning_rate": 3.2442316686675783e-06,
      "loss": 0.4437,
      "step": 1380
    },
    {
      "epoch": 1.9751243781094527,
      "grad_norm": 0.595196616315948,
      "learning_rate": 3.1670665359532415e-06,
      "loss": 0.4438,
      "step": 1390
    },
    {
      "epoch": 1.9893390191897655,
      "grad_norm": 0.5920175988784884,
      "learning_rate": 3.090402508946249e-06,
      "loss": 0.4373,
      "step": 1400
    },
    {
      "epoch": 2.0042643923240937,
      "grad_norm": 0.5358117816177491,
      "learning_rate": 3.0142605468260976e-06,
      "loss": 0.4572,
      "step": 1410
    },
    {
      "epoch": 2.0184790334044065,
      "grad_norm": 0.5576529014623839,
      "learning_rate": 2.9386614660449598e-06,
      "loss": 0.4263,
      "step": 1420
    },
    {
      "epoch": 2.0326936744847193,
      "grad_norm": 0.5822207756082265,
      "learning_rate": 2.8636259346366666e-06,
      "loss": 0.4385,
      "step": 1430
    },
    {
      "epoch": 2.046908315565032,
      "grad_norm": 0.5829696226017161,
      "learning_rate": 2.7891744665662824e-06,
      "loss": 0.4218,
      "step": 1440
    },
    {
      "epoch": 2.061122956645345,
      "grad_norm": 0.581038802974167,
      "learning_rate": 2.7153274161217847e-06,
      "loss": 0.4404,
      "step": 1450
    },
    {
      "epoch": 2.0753375977256576,
      "grad_norm": 0.5347733423294494,
      "learning_rate": 2.642104972349403e-06,
      "loss": 0.4198,
      "step": 1460
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 0.543588717662816,
      "learning_rate": 2.5695271535341443e-06,
      "loss": 0.4411,
      "step": 1470
    },
    {
      "epoch": 2.1037668798862827,
      "grad_norm": 0.5720716950974386,
      "learning_rate": 2.4976138017269906e-06,
      "loss": 0.4412,
      "step": 1480
    },
    {
      "epoch": 2.1179815209665955,
      "grad_norm": 0.5603303814513405,
      "learning_rate": 2.4263845773202738e-06,
      "loss": 0.4298,
      "step": 1490
    },
    {
      "epoch": 2.1321961620469083,
      "grad_norm": 0.5598020164471934,
      "learning_rate": 2.355858953672728e-06,
      "loss": 0.4397,
      "step": 1500
    },
    {
      "epoch": 2.1321961620469083,
      "eval_loss": 0.5144844651222229,
      "eval_runtime": 231.3372,
      "eval_samples_per_second": 21.613,
      "eval_steps_per_second": 2.702,
      "step": 1500
    },
    {
      "epoch": 2.146410803127221,
      "grad_norm": 0.552503902368828,
      "learning_rate": 2.286056211785665e-06,
      "loss": 0.4339,
      "step": 1510
    },
    {
      "epoch": 2.160625444207534,
      "grad_norm": 0.5606006250953287,
      "learning_rate": 2.2169954350317372e-06,
      "loss": 0.4408,
      "step": 1520
    },
    {
      "epoch": 2.1748400852878467,
      "grad_norm": 0.5478812953721691,
      "learning_rate": 2.148695503937745e-06,
      "loss": 0.4387,
      "step": 1530
    },
    {
      "epoch": 2.189054726368159,
      "grad_norm": 0.5327946624830445,
      "learning_rate": 2.081175091022877e-06,
      "loss": 0.4395,
      "step": 1540
    },
    {
      "epoch": 2.203269367448472,
      "grad_norm": 0.5793441763139854,
      "learning_rate": 2.014452655693839e-06,
      "loss": 0.4394,
      "step": 1550
    },
    {
      "epoch": 2.2174840085287846,
      "grad_norm": 0.5656062787295659,
      "learning_rate": 1.9485464391982282e-06,
      "loss": 0.4357,
      "step": 1560
    },
    {
      "epoch": 2.2316986496090974,
      "grad_norm": 0.575605417624959,
      "learning_rate": 1.8834744596375664e-06,
      "loss": 0.4364,
      "step": 1570
    },
    {
      "epoch": 2.24591329068941,
      "grad_norm": 0.5511255544375402,
      "learning_rate": 1.8192545070413281e-06,
      "loss": 0.4278,
      "step": 1580
    },
    {
      "epoch": 2.260127931769723,
      "grad_norm": 0.5527437049458982,
      "learning_rate": 1.755904138503316e-06,
      "loss": 0.4281,
      "step": 1590
    },
    {
      "epoch": 2.2743425728500357,
      "grad_norm": 0.5567466633261947,
      "learning_rate": 1.6934406733817417e-06,
      "loss": 0.4326,
      "step": 1600
    },
    {
      "epoch": 2.288557213930348,
      "grad_norm": 0.5542280645431442,
      "learning_rate": 1.6318811885642749e-06,
      "loss": 0.4426,
      "step": 1610
    },
    {
      "epoch": 2.302771855010661,
      "grad_norm": 0.5385541713578466,
      "learning_rate": 1.5712425137993976e-06,
      "loss": 0.4363,
      "step": 1620
    },
    {
      "epoch": 2.3169864960909736,
      "grad_norm": 0.5568902806764913,
      "learning_rate": 1.5115412270953166e-06,
      "loss": 0.4346,
      "step": 1630
    },
    {
      "epoch": 2.3312011371712864,
      "grad_norm": 0.6058817336949707,
      "learning_rate": 1.4527936501877033e-06,
      "loss": 0.4321,
      "step": 1640
    },
    {
      "epoch": 2.345415778251599,
      "grad_norm": 0.5318061336854996,
      "learning_rate": 1.3950158440774958e-06,
      "loss": 0.418,
      "step": 1650
    },
    {
      "epoch": 2.359630419331912,
      "grad_norm": 0.5553934307354601,
      "learning_rate": 1.3382236046399722e-06,
      "loss": 0.4143,
      "step": 1660
    },
    {
      "epoch": 2.3738450604122248,
      "grad_norm": 0.5429114693602697,
      "learning_rate": 1.2824324583063303e-06,
      "loss": 0.4413,
      "step": 1670
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 0.5286255307861315,
      "learning_rate": 1.2276576578189065e-06,
      "loss": 0.4218,
      "step": 1680
    },
    {
      "epoch": 2.40227434257285,
      "grad_norm": 0.5513151927730859,
      "learning_rate": 1.1739141780612306e-06,
      "loss": 0.4278,
      "step": 1690
    },
    {
      "epoch": 2.4164889836531627,
      "grad_norm": 0.5523983447772243,
      "learning_rate": 1.1212167119640439e-06,
      "loss": 0.4364,
      "step": 1700
    },
    {
      "epoch": 2.4307036247334755,
      "grad_norm": 0.5707718364766191,
      "learning_rate": 1.069579666488395e-06,
      "loss": 0.429,
      "step": 1710
    },
    {
      "epoch": 2.4449182658137882,
      "grad_norm": 0.5620343463380818,
      "learning_rate": 1.0190171586869258e-06,
      "loss": 0.4239,
      "step": 1720
    },
    {
      "epoch": 2.459132906894101,
      "grad_norm": 0.5335024558606225,
      "learning_rate": 9.695430118444049e-07,
      "loss": 0.4199,
      "step": 1730
    },
    {
      "epoch": 2.473347547974414,
      "grad_norm": 0.5088660962728315,
      "learning_rate": 9.21170751698583e-07,
      "loss": 0.4294,
      "step": 1740
    },
    {
      "epoch": 2.487562189054726,
      "grad_norm": 0.5854213827338935,
      "learning_rate": 8.739136027423894e-07,
      "loss": 0.4407,
      "step": 1750
    },
    {
      "epoch": 2.501776830135039,
      "grad_norm": 0.570296065993969,
      "learning_rate": 8.277844846084898e-07,
      "loss": 0.4273,
      "step": 1760
    },
    {
      "epoch": 2.5159914712153517,
      "grad_norm": 0.5432303322927625,
      "learning_rate": 7.827960085371855e-07,
      "loss": 0.4372,
      "step": 1770
    },
    {
      "epoch": 2.5302061122956645,
      "grad_norm": 0.5462735332010124,
      "learning_rate": 7.389604739286271e-07,
      "loss": 0.4384,
      "step": 1780
    },
    {
      "epoch": 2.5444207533759773,
      "grad_norm": 0.5514867453544212,
      "learning_rate": 6.962898649802824e-07,
      "loss": 0.4279,
      "step": 1790
    },
    {
      "epoch": 2.55863539445629,
      "grad_norm": 0.5374953523760396,
      "learning_rate": 6.547958474105726e-07,
      "loss": 0.4207,
      "step": 1800
    },
    {
      "epoch": 2.572850035536603,
      "grad_norm": 0.5368534822557964,
      "learning_rate": 6.144897652695864e-07,
      "loss": 0.4425,
      "step": 1810
    },
    {
      "epoch": 2.587064676616915,
      "grad_norm": 0.5331851936220864,
      "learning_rate": 5.753826378377287e-07,
      "loss": 0.4227,
      "step": 1820
    },
    {
      "epoch": 2.6012793176972284,
      "grad_norm": 0.532292119043563,
      "learning_rate": 5.374851566131561e-07,
      "loss": 0.4332,
      "step": 1830
    },
    {
      "epoch": 2.6154939587775408,
      "grad_norm": 0.522004918322321,
      "learning_rate": 5.008076823888319e-07,
      "loss": 0.4363,
      "step": 1840
    },
    {
      "epoch": 2.6297085998578535,
      "grad_norm": 0.5747875880967467,
      "learning_rate": 4.653602424199877e-07,
      "loss": 0.441,
      "step": 1850
    },
    {
      "epoch": 2.6439232409381663,
      "grad_norm": 0.5701129333105889,
      "learning_rate": 4.3115252768276827e-07,
      "loss": 0.4281,
      "step": 1860
    },
    {
      "epoch": 2.658137882018479,
      "grad_norm": 0.5171567622519524,
      "learning_rate": 3.9819389022482226e-07,
      "loss": 0.439,
      "step": 1870
    },
    {
      "epoch": 2.672352523098792,
      "grad_norm": 0.5506453302407677,
      "learning_rate": 3.6649334060854027e-07,
      "loss": 0.4261,
      "step": 1880
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 0.5611217928624682,
      "learning_rate": 3.360595454476595e-07,
      "loss": 0.4296,
      "step": 1890
    },
    {
      "epoch": 2.7007818052594175,
      "grad_norm": 0.5693324256729151,
      "learning_rate": 3.069008250378974e-07,
      "loss": 0.4349,
      "step": 1900
    },
    {
      "epoch": 2.71499644633973,
      "grad_norm": 0.5448625830798527,
      "learning_rate": 2.7902515108226613e-07,
      "loss": 0.4167,
      "step": 1910
    },
    {
      "epoch": 2.7292110874200426,
      "grad_norm": 0.5178772865930678,
      "learning_rate": 2.5244014451168863e-07,
      "loss": 0.4335,
      "step": 1920
    },
    {
      "epoch": 2.7434257285003554,
      "grad_norm": 0.555692357887426,
      "learning_rate": 2.271530734015104e-07,
      "loss": 0.435,
      "step": 1930
    },
    {
      "epoch": 2.757640369580668,
      "grad_norm": 0.5575021836492726,
      "learning_rate": 2.0317085098448373e-07,
      "loss": 0.4222,
      "step": 1940
    },
    {
      "epoch": 2.771855010660981,
      "grad_norm": 0.5354458107562179,
      "learning_rate": 1.8050003376075708e-07,
      "loss": 0.4426,
      "step": 1950
    },
    {
      "epoch": 2.7860696517412933,
      "grad_norm": 0.5182277398666688,
      "learning_rate": 1.591468197053919e-07,
      "loss": 0.4359,
      "step": 1960
    },
    {
      "epoch": 2.8002842928216065,
      "grad_norm": 0.5422707453993397,
      "learning_rate": 1.3911704657390113e-07,
      "loss": 0.4243,
      "step": 1970
    },
    {
      "epoch": 2.814498933901919,
      "grad_norm": 0.5242061992724598,
      "learning_rate": 1.2041619030626283e-07,
      "loss": 0.4323,
      "step": 1980
    },
    {
      "epoch": 2.8287135749822316,
      "grad_norm": 0.5113276517453622,
      "learning_rate": 1.0304936352985351e-07,
      "loss": 0.4273,
      "step": 1990
    },
    {
      "epoch": 2.8429282160625444,
      "grad_norm": 0.5355864223633939,
      "learning_rate": 8.702131416170657e-08,
      "loss": 0.436,
      "step": 2000
    },
    {
      "epoch": 2.8429282160625444,
      "eval_loss": 0.5102300047874451,
      "eval_runtime": 231.9212,
      "eval_samples_per_second": 21.559,
      "eval_steps_per_second": 2.695,
      "step": 2000
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.5388940284634288,
      "learning_rate": 7.233642411048014e-08,
      "loss": 0.437,
      "step": 2010
    },
    {
      "epoch": 2.87135749822317,
      "grad_norm": 0.5557321412758267,
      "learning_rate": 5.899870807848762e-08,
      "loss": 0.4304,
      "step": 2020
    },
    {
      "epoch": 2.8855721393034823,
      "grad_norm": 0.5448610951969772,
      "learning_rate": 4.701181246411501e-08,
      "loss": 0.4275,
      "step": 2030
    },
    {
      "epoch": 2.8997867803837956,
      "grad_norm": 0.5737836534540497,
      "learning_rate": 3.6379014364935075e-08,
      "loss": 0.4349,
      "step": 2040
    },
    {
      "epoch": 2.914001421464108,
      "grad_norm": 0.55622828041129,
      "learning_rate": 2.7103220681780616e-08,
      "loss": 0.4291,
      "step": 2050
    },
    {
      "epoch": 2.9282160625444207,
      "grad_norm": 0.5234759202957545,
      "learning_rate": 1.9186967324026364e-08,
      "loss": 0.4283,
      "step": 2060
    },
    {
      "epoch": 2.9424307036247335,
      "grad_norm": 0.533033568944861,
      "learning_rate": 1.2632418516296263e-08,
      "loss": 0.4269,
      "step": 2070
    },
    {
      "epoch": 2.9566453447050463,
      "grad_norm": 0.5413416159064122,
      "learning_rate": 7.4413662067884806e-09,
      "loss": 0.4291,
      "step": 2080
    },
    {
      "epoch": 2.970859985785359,
      "grad_norm": 0.5523224114708861,
      "learning_rate": 3.615229577371149e-09,
      "loss": 0.432,
      "step": 2090
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 0.5328203745561761,
      "learning_rate": 1.1550546555960662e-09,
      "loss": 0.4398,
      "step": 2100
    },
    {
      "epoch": 2.9992892679459846,
      "grad_norm": 0.5266787263911724,
      "learning_rate": 6.151402872134337e-11,
      "loss": 0.422,
      "step": 2110
    },
    {
      "epoch": 3.0,
      "step": 2111,
      "total_flos": 1256820975730688.0,
      "train_loss": 0.22985298537010831,
      "train_runtime": 4731.3341,
      "train_samples_per_second": 28.533,
      "train_steps_per_second": 0.446
    }
  ],
  "logging_steps": 10,
  "max_steps": 2112,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1256820975730688.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}