{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.6255767963085037,
  "eval_steps": 500,
  "global_step": 5500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0065919578114700065,
      "grad_norm": 10.419755935668945,
      "learning_rate": 7.017543859649123e-06,
      "loss": 1.3206,
      "step": 10
    },
    {
      "epoch": 0.013183915622940013,
      "grad_norm": 7.308764457702637,
      "learning_rate": 1.4035087719298246e-05,
      "loss": 1.0588,
      "step": 20
    },
    {
      "epoch": 0.01977587343441002,
      "grad_norm": 3.7246360778808594,
      "learning_rate": 2.105263157894737e-05,
      "loss": 0.4529,
      "step": 30
    },
    {
      "epoch": 0.026367831245880026,
      "grad_norm": 1.944443702697754,
      "learning_rate": 2.8070175438596492e-05,
      "loss": 0.3158,
      "step": 40
    },
    {
      "epoch": 0.03295978905735003,
      "grad_norm": 1.4747883081436157,
      "learning_rate": 3.508771929824561e-05,
      "loss": 0.2119,
      "step": 50
    },
    {
      "epoch": 0.03955174686882004,
      "grad_norm": 1.336323857307434,
      "learning_rate": 4.210526315789474e-05,
      "loss": 0.1738,
      "step": 60
    },
    {
      "epoch": 0.04614370468029005,
      "grad_norm": 1.3995416164398193,
      "learning_rate": 4.912280701754386e-05,
      "loss": 0.1525,
      "step": 70
    },
    {
      "epoch": 0.05273566249176005,
      "grad_norm": 1.2422436475753784,
      "learning_rate": 5.6140350877192984e-05,
      "loss": 0.1339,
      "step": 80
    },
    {
      "epoch": 0.05932762030323006,
      "grad_norm": 1.134950876235962,
      "learning_rate": 6.31578947368421e-05,
      "loss": 0.1051,
      "step": 90
    },
    {
      "epoch": 0.06591957811470006,
      "grad_norm": 1.2677438259124756,
      "learning_rate": 7.017543859649122e-05,
      "loss": 0.1054,
      "step": 100
    },
    {
      "epoch": 0.07251153592617007,
      "grad_norm": 0.9868601560592651,
      "learning_rate": 7.719298245614036e-05,
      "loss": 0.0977,
      "step": 110
    },
    {
      "epoch": 0.07910349373764008,
      "grad_norm": 1.3295618295669556,
      "learning_rate": 8.421052631578948e-05,
      "loss": 0.0948,
      "step": 120
    },
    {
      "epoch": 0.08569545154911008,
      "grad_norm": 0.7764300107955933,
      "learning_rate": 9.12280701754386e-05,
      "loss": 0.0714,
      "step": 130
    },
    {
      "epoch": 0.0922874093605801,
      "grad_norm": 1.17441987991333,
      "learning_rate": 9.824561403508771e-05,
      "loss": 0.0742,
      "step": 140
    },
    {
      "epoch": 0.09887936717205009,
      "grad_norm": 1.4673572778701782,
      "learning_rate": 0.00010526315789473685,
      "loss": 0.0782,
      "step": 150
    },
    {
      "epoch": 0.1054713249835201,
      "grad_norm": 0.6098851561546326,
      "learning_rate": 0.00011228070175438597,
      "loss": 0.0778,
      "step": 160
    },
    {
      "epoch": 0.11206328279499012,
      "grad_norm": 0.921156644821167,
      "learning_rate": 0.00011929824561403509,
      "loss": 0.0761,
      "step": 170
    },
    {
      "epoch": 0.11865524060646011,
      "grad_norm": 0.9673042297363281,
      "learning_rate": 0.0001263157894736842,
      "loss": 0.0723,
      "step": 180
    },
    {
      "epoch": 0.12524719841793014,
      "grad_norm": 0.6645211577415466,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.0672,
      "step": 190
    },
    {
      "epoch": 0.13183915622940012,
      "grad_norm": 1.5125294923782349,
      "learning_rate": 0.00014035087719298245,
      "loss": 0.0778,
      "step": 200
    },
    {
      "epoch": 0.13843111404087013,
      "grad_norm": 0.8468068242073059,
      "learning_rate": 0.00014736842105263158,
      "loss": 0.0674,
      "step": 210
    },
    {
      "epoch": 0.14502307185234015,
      "grad_norm": 0.9815834164619446,
      "learning_rate": 0.0001543859649122807,
      "loss": 0.068,
      "step": 220
    },
    {
      "epoch": 0.15161502966381016,
      "grad_norm": 0.7477302551269531,
      "learning_rate": 0.00016140350877192982,
      "loss": 0.0611,
      "step": 230
    },
    {
      "epoch": 0.15820698747528017,
      "grad_norm": 0.815642237663269,
      "learning_rate": 0.00016842105263157895,
      "loss": 0.0631,
      "step": 240
    },
    {
      "epoch": 0.16479894528675015,
      "grad_norm": 0.5081785321235657,
      "learning_rate": 0.00017543859649122806,
      "loss": 0.0607,
      "step": 250
    },
    {
      "epoch": 0.17139090309822017,
      "grad_norm": 0.718622624874115,
      "learning_rate": 0.0001824561403508772,
      "loss": 0.063,
      "step": 260
    },
    {
      "epoch": 0.17798286090969018,
      "grad_norm": 0.6826403141021729,
      "learning_rate": 0.00018947368421052632,
      "loss": 0.0611,
      "step": 270
    },
    {
      "epoch": 0.1845748187211602,
      "grad_norm": 0.730829656124115,
      "learning_rate": 0.00019649122807017543,
      "loss": 0.0636,
      "step": 280
    },
    {
      "epoch": 0.1911667765326302,
      "grad_norm": 0.976826012134552,
      "learning_rate": 0.00019999957723376656,
      "loss": 0.05,
      "step": 290
    },
    {
      "epoch": 0.19775873434410018,
      "grad_norm": 0.943250298500061,
      "learning_rate": 0.00019999619512534684,
      "loss": 0.0578,
      "step": 300
    },
    {
      "epoch": 0.2043506921555702,
      "grad_norm": 0.8867354393005371,
      "learning_rate": 0.00019998943102289446,
      "loss": 0.0623,
      "step": 310
    },
    {
      "epoch": 0.2109426499670402,
      "grad_norm": 0.4728466272354126,
      "learning_rate": 0.0001999792851551796,
      "loss": 0.0527,
      "step": 320
    },
    {
      "epoch": 0.21753460777851022,
      "grad_norm": 1.1633834838867188,
      "learning_rate": 0.00019996575786534806,
      "loss": 0.0568,
      "step": 330
    },
    {
      "epoch": 0.22412656558998023,
      "grad_norm": 1.2754548788070679,
      "learning_rate": 0.00019994884961090933,
      "loss": 0.0542,
      "step": 340
    },
    {
      "epoch": 0.23071852340145024,
      "grad_norm": 1.3917484283447266,
      "learning_rate": 0.0001999285609637213,
      "loss": 0.0627,
      "step": 350
    },
    {
      "epoch": 0.23731048121292023,
      "grad_norm": 0.7138040065765381,
      "learning_rate": 0.00019990489260997095,
      "loss": 0.0548,
      "step": 360
    },
    {
      "epoch": 0.24390243902439024,
      "grad_norm": 0.48101598024368286,
      "learning_rate": 0.00019987784535015107,
      "loss": 0.039,
      "step": 370
    },
    {
      "epoch": 0.2504943968358603,
      "grad_norm": 0.6185160279273987,
      "learning_rate": 0.00019984742009903313,
      "loss": 0.0516,
      "step": 380
    },
    {
      "epoch": 0.25708635464733026,
      "grad_norm": 0.5346279144287109,
      "learning_rate": 0.00019981361788563647,
      "loss": 0.0594,
      "step": 390
    },
    {
      "epoch": 0.26367831245880025,
      "grad_norm": 0.644879937171936,
      "learning_rate": 0.00019977643985319345,
      "loss": 0.0415,
      "step": 400
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 0.8102220296859741,
      "learning_rate": 0.00019973588725911076,
      "loss": 0.0466,
      "step": 410
    },
    {
      "epoch": 0.27686222808174027,
      "grad_norm": 0.6942636370658875,
      "learning_rate": 0.00019969196147492685,
      "loss": 0.0498,
      "step": 420
    },
    {
      "epoch": 0.2834541858932103,
      "grad_norm": 0.9616207480430603,
      "learning_rate": 0.0001996446639862657,
      "loss": 0.0489,
      "step": 430
    },
    {
      "epoch": 0.2900461437046803,
      "grad_norm": 0.7776756882667542,
      "learning_rate": 0.00019959399639278636,
      "loss": 0.0481,
      "step": 440
    },
    {
      "epoch": 0.2966381015161503,
      "grad_norm": 1.2910106182098389,
      "learning_rate": 0.0001995399604081291,
      "loss": 0.045,
      "step": 450
    },
    {
      "epoch": 0.3032300593276203,
      "grad_norm": 0.5743927955627441,
      "learning_rate": 0.00019948255785985717,
      "loss": 0.0419,
      "step": 460
    },
    {
      "epoch": 0.3098220171390903,
      "grad_norm": 0.6184401512145996,
      "learning_rate": 0.0001994217906893952,
      "loss": 0.0445,
      "step": 470
    },
    {
      "epoch": 0.31641397495056034,
      "grad_norm": 0.7128023505210876,
      "learning_rate": 0.00019935766095196348,
      "loss": 0.038,
      "step": 480
    },
    {
      "epoch": 0.3230059327620303,
      "grad_norm": 0.5777243375778198,
      "learning_rate": 0.00019929017081650845,
      "loss": 0.0417,
      "step": 490
    },
    {
      "epoch": 0.3295978905735003,
      "grad_norm": 0.6238442659378052,
      "learning_rate": 0.00019921932256562928,
      "loss": 0.0343,
      "step": 500
    },
    {
      "epoch": 0.33618984838497035,
      "grad_norm": 0.5910335183143616,
      "learning_rate": 0.00019914511859550076,
      "loss": 0.0428,
      "step": 510
    },
    {
      "epoch": 0.34278180619644033,
      "grad_norm": 0.5180881023406982,
      "learning_rate": 0.00019906756141579226,
      "loss": 0.034,
      "step": 520
    },
    {
      "epoch": 0.34937376400791037,
      "grad_norm": 0.34267082810401917,
      "learning_rate": 0.00019898665364958274,
      "loss": 0.0391,
      "step": 530
    },
    {
      "epoch": 0.35596572181938035,
      "grad_norm": 0.4261837303638458,
      "learning_rate": 0.00019890239803327213,
      "loss": 0.0453,
      "step": 540
    },
    {
      "epoch": 0.36255767963085034,
      "grad_norm": 0.5537737011909485,
      "learning_rate": 0.0001988147974164888,
      "loss": 0.0323,
      "step": 550
    },
    {
      "epoch": 0.3691496374423204,
      "grad_norm": 0.419604629278183,
      "learning_rate": 0.0001987238547619931,
      "loss": 0.037,
      "step": 560
    },
    {
      "epoch": 0.37574159525379036,
      "grad_norm": 0.8267730474472046,
      "learning_rate": 0.00019862957314557722,
      "loss": 0.0452,
      "step": 570
    },
    {
      "epoch": 0.3823335530652604,
      "grad_norm": 0.4021684229373932,
      "learning_rate": 0.0001985319557559611,
      "loss": 0.042,
      "step": 580
    },
    {
      "epoch": 0.3889255108767304,
      "grad_norm": 0.5992162227630615,
      "learning_rate": 0.0001984310058946847,
      "loss": 0.0323,
      "step": 590
    },
    {
      "epoch": 0.39551746868820037,
      "grad_norm": 0.5908779501914978,
      "learning_rate": 0.00019832672697599616,
      "loss": 0.0391,
      "step": 600
    },
    {
      "epoch": 0.4021094264996704,
      "grad_norm": 0.6762053370475769,
      "learning_rate": 0.0001982191225267365,
      "loss": 0.0444,
      "step": 610
    },
    {
      "epoch": 0.4087013843111404,
      "grad_norm": 0.6451637148857117,
      "learning_rate": 0.00019810819618622025,
      "loss": 0.0405,
      "step": 620
    },
    {
      "epoch": 0.41529334212261043,
      "grad_norm": 0.40719446539878845,
      "learning_rate": 0.00019799395170611237,
      "loss": 0.032,
      "step": 630
    },
    {
      "epoch": 0.4218852999340804,
      "grad_norm": 0.3678116202354431,
      "learning_rate": 0.00019787639295030135,
      "loss": 0.0295,
      "step": 640
    },
    {
      "epoch": 0.42847725774555045,
      "grad_norm": 0.5194733738899231,
      "learning_rate": 0.00019775552389476864,
      "loss": 0.0336,
      "step": 650
    },
    {
      "epoch": 0.43506921555702044,
      "grad_norm": 0.3977694511413574,
      "learning_rate": 0.00019763134862745395,
      "loss": 0.0339,
      "step": 660
    },
    {
      "epoch": 0.4416611733684904,
      "grad_norm": 0.4669642150402069,
      "learning_rate": 0.00019750387134811725,
      "loss": 0.0256,
      "step": 670
    },
    {
      "epoch": 0.44825313117996046,
      "grad_norm": 0.36751696467399597,
      "learning_rate": 0.00019737309636819658,
      "loss": 0.0437,
      "step": 680
    },
    {
      "epoch": 0.45484508899143045,
      "grad_norm": 0.8601805567741394,
      "learning_rate": 0.00019723902811066222,
      "loss": 0.0311,
      "step": 690
    },
    {
      "epoch": 0.4614370468029005,
      "grad_norm": 0.3772714138031006,
      "learning_rate": 0.00019710167110986718,
      "loss": 0.0334,
      "step": 700
    },
    {
      "epoch": 0.46802900461437047,
      "grad_norm": 0.5447872281074524,
      "learning_rate": 0.00019696103001139373,
      "loss": 0.0379,
      "step": 710
    },
    {
      "epoch": 0.47462096242584045,
      "grad_norm": 0.3882730007171631,
      "learning_rate": 0.00019681710957189651,
      "loss": 0.0314,
      "step": 720
    },
    {
      "epoch": 0.4812129202373105,
      "grad_norm": 0.3971300423145294,
      "learning_rate": 0.00019666991465894134,
      "loss": 0.0349,
      "step": 730
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.41290801763534546,
      "learning_rate": 0.00019651945025084082,
      "loss": 0.026,
      "step": 740
    },
    {
      "epoch": 0.4943968358602505,
      "grad_norm": 0.5641320943832397,
      "learning_rate": 0.00019636572143648588,
      "loss": 0.0415,
      "step": 750
    },
    {
      "epoch": 0.5009887936717206,
      "grad_norm": 0.537247896194458,
      "learning_rate": 0.00019620873341517366,
      "loss": 0.0336,
      "step": 760
    },
    {
      "epoch": 0.5075807514831905,
      "grad_norm": 0.3074246644973755,
      "learning_rate": 0.00019604849149643173,
      "loss": 0.0353,
      "step": 770
    },
    {
      "epoch": 0.5141727092946605,
      "grad_norm": 0.5712188482284546,
      "learning_rate": 0.0001958850010998384,
      "loss": 0.0356,
      "step": 780
    },
    {
      "epoch": 0.5207646671061306,
      "grad_norm": 0.7442903518676758,
      "learning_rate": 0.00019571826775483954,
      "loss": 0.0329,
      "step": 790
    },
    {
      "epoch": 0.5273566249176005,
      "grad_norm": 0.3076368570327759,
      "learning_rate": 0.00019554829710056147,
      "loss": 0.0366,
      "step": 800
    },
    {
      "epoch": 0.5339485827290705,
      "grad_norm": 0.4598945081233978,
      "learning_rate": 0.00019537509488562032,
      "loss": 0.0271,
      "step": 810
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 0.6214160919189453,
      "learning_rate": 0.0001951986669679275,
      "loss": 0.0414,
      "step": 820
    },
    {
      "epoch": 0.5471324983520105,
      "grad_norm": 0.5321261882781982,
      "learning_rate": 0.00019501901931449176,
      "loss": 0.0322,
      "step": 830
    },
    {
      "epoch": 0.5537244561634805,
      "grad_norm": 0.5436400771141052,
      "learning_rate": 0.00019483615800121716,
      "loss": 0.0261,
      "step": 840
    },
    {
      "epoch": 0.5603164139749506,
      "grad_norm": 0.39354392886161804,
      "learning_rate": 0.00019465008921269774,
      "loss": 0.0299,
      "step": 850
    },
    {
      "epoch": 0.5669083717864206,
      "grad_norm": 0.5768340229988098,
      "learning_rate": 0.00019446081924200825,
      "loss": 0.034,
      "step": 860
    },
    {
      "epoch": 0.5735003295978905,
      "grad_norm": 0.43069860339164734,
      "learning_rate": 0.0001942683544904914,
      "loss": 0.0286,
      "step": 870
    },
    {
      "epoch": 0.5800922874093606,
      "grad_norm": 0.6545283794403076,
      "learning_rate": 0.0001940727014675412,
      "loss": 0.0353,
      "step": 880
    },
    {
      "epoch": 0.5866842452208306,
      "grad_norm": 0.4092136323451996,
      "learning_rate": 0.00019387386679038307,
      "loss": 0.0284,
      "step": 890
    },
    {
      "epoch": 0.5932762030323006,
      "grad_norm": 0.38882556557655334,
      "learning_rate": 0.0001936718571838498,
      "loss": 0.0294,
      "step": 900
    },
    {
      "epoch": 0.5998681608437706,
      "grad_norm": 0.6306661367416382,
      "learning_rate": 0.0001934666794801541,
      "loss": 0.0331,
      "step": 910
    },
    {
      "epoch": 0.6064601186552406,
      "grad_norm": 0.35673242807388306,
      "learning_rate": 0.00019325834061865776,
      "loss": 0.0356,
      "step": 920
    },
    {
      "epoch": 0.6130520764667106,
      "grad_norm": 0.5133143067359924,
      "learning_rate": 0.00019304684764563675,
      "loss": 0.0349,
      "step": 930
    },
    {
      "epoch": 0.6196440342781806,
      "grad_norm": 0.32806074619293213,
      "learning_rate": 0.00019283220771404297,
      "loss": 0.0333,
      "step": 940
    },
    {
      "epoch": 0.6262359920896506,
      "grad_norm": 0.4773825705051422,
      "learning_rate": 0.00019261442808326225,
      "loss": 0.0302,
      "step": 950
    },
    {
      "epoch": 0.6328279499011207,
      "grad_norm": 0.7042701244354248,
      "learning_rate": 0.000192393516118869,
      "loss": 0.0375,
      "step": 960
    },
    {
      "epoch": 0.6394199077125906,
      "grad_norm": 0.40759798884391785,
      "learning_rate": 0.00019216947929237696,
      "loss": 0.0273,
      "step": 970
    },
    {
      "epoch": 0.6460118655240606,
      "grad_norm": 0.2918468713760376,
      "learning_rate": 0.0001919423251809865,
      "loss": 0.0321,
      "step": 980
    },
    {
      "epoch": 0.6526038233355307,
      "grad_norm": 0.3963848352432251,
      "learning_rate": 0.00019171206146732848,
      "loss": 0.0272,
      "step": 990
    },
    {
      "epoch": 0.6591957811470006,
      "grad_norm": 0.4451894462108612,
      "learning_rate": 0.00019147869593920424,
      "loss": 0.0348,
      "step": 1000
    },
    {
      "epoch": 0.6657877389584707,
      "grad_norm": 0.5182775855064392,
      "learning_rate": 0.00019124223648932232,
      "loss": 0.037,
      "step": 1010
    },
    {
      "epoch": 0.6723796967699407,
      "grad_norm": 0.6909258961677551,
      "learning_rate": 0.00019100269111503148,
      "loss": 0.0307,
      "step": 1020
    },
    {
      "epoch": 0.6789716545814107,
      "grad_norm": 0.4598976969718933,
      "learning_rate": 0.0001907600679180503,
      "loss": 0.0308,
      "step": 1030
    },
    {
      "epoch": 0.6855636123928807,
      "grad_norm": 0.5611470937728882,
      "learning_rate": 0.00019051437510419296,
      "loss": 0.0303,
      "step": 1040
    },
    {
      "epoch": 0.6921555702043507,
      "grad_norm": 0.2982472777366638,
      "learning_rate": 0.00019026562098309202,
      "loss": 0.0273,
      "step": 1050
    },
    {
      "epoch": 0.6987475280158207,
      "grad_norm": 0.4202234745025635,
      "learning_rate": 0.00019001381396791707,
      "loss": 0.0348,
      "step": 1060
    },
    {
      "epoch": 0.7053394858272907,
      "grad_norm": 0.3558894991874695,
      "learning_rate": 0.0001897589625750904,
      "loss": 0.0233,
      "step": 1070
    },
    {
      "epoch": 0.7119314436387607,
      "grad_norm": 0.4467407464981079,
      "learning_rate": 0.00018950107542399885,
      "loss": 0.0284,
      "step": 1080
    },
    {
      "epoch": 0.7185234014502307,
      "grad_norm": 1.036746859550476,
      "learning_rate": 0.00018924016123670235,
      "loss": 0.0969,
      "step": 1090
    },
    {
      "epoch": 0.7251153592617007,
      "grad_norm": 0.5457211136817932,
      "learning_rate": 0.00018897622883763888,
      "loss": 0.0622,
      "step": 1100
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 0.675666093826294,
      "learning_rate": 0.0001887092871533261,
      "loss": 0.0585,
      "step": 1110
    },
    {
      "epoch": 0.7382992748846408,
      "grad_norm": 1.0533791780471802,
      "learning_rate": 0.0001884393452120594,
      "loss": 0.0619,
      "step": 1120
    },
    {
      "epoch": 0.7448912326961108,
      "grad_norm": 0.5425913333892822,
      "learning_rate": 0.0001881664121436064,
      "loss": 0.0619,
      "step": 1130
    },
    {
      "epoch": 0.7514831905075807,
      "grad_norm": 0.46993035078048706,
      "learning_rate": 0.0001878904971788985,
      "loss": 0.0614,
      "step": 1140
    },
    {
      "epoch": 0.7580751483190508,
      "grad_norm": 0.5931456089019775,
      "learning_rate": 0.00018761160964971837,
      "loss": 0.0636,
      "step": 1150
    },
    {
      "epoch": 0.7646671061305208,
      "grad_norm": 0.42014068365097046,
      "learning_rate": 0.00018732975898838445,
      "loss": 0.0637,
      "step": 1160
    },
    {
      "epoch": 0.7712590639419907,
      "grad_norm": 0.5799768567085266,
      "learning_rate": 0.00018704495472743207,
      "loss": 0.062,
      "step": 1170
    },
    {
      "epoch": 0.7778510217534608,
      "grad_norm": 0.9316193461418152,
      "learning_rate": 0.0001867572064992908,
      "loss": 0.0674,
      "step": 1180
    },
    {
      "epoch": 0.7844429795649308,
      "grad_norm": 0.9099822044372559,
      "learning_rate": 0.00018646652403595885,
      "loss": 0.0656,
      "step": 1190
    },
    {
      "epoch": 0.7910349373764007,
      "grad_norm": 0.5124749541282654,
      "learning_rate": 0.00018617291716867382,
      "loss": 0.0801,
      "step": 1200
    },
    {
      "epoch": 0.7976268951878708,
      "grad_norm": 0.6862666010856628,
      "learning_rate": 0.00018587639582758031,
      "loss": 0.0668,
      "step": 1210
    },
    {
      "epoch": 0.8042188529993408,
      "grad_norm": 0.6310227513313293,
      "learning_rate": 0.00018557697004139393,
      "loss": 0.0663,
      "step": 1220
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 0.6013323068618774,
      "learning_rate": 0.00018527464993706226,
      "loss": 0.065,
      "step": 1230
    },
    {
      "epoch": 0.8174027686222808,
      "grad_norm": 0.5435302257537842,
      "learning_rate": 0.00018496944573942218,
      "loss": 0.0569,
      "step": 1240
    },
    {
      "epoch": 0.8239947264337508,
      "grad_norm": 0.6998788118362427,
      "learning_rate": 0.0001846613677708542,
      "loss": 0.0656,
      "step": 1250
    },
    {
      "epoch": 0.8305866842452209,
      "grad_norm": 0.6783934235572815,
      "learning_rate": 0.0001843504264509333,
      "loss": 0.0524,
      "step": 1260
    },
    {
      "epoch": 0.8371786420566908,
      "grad_norm": 0.6428571343421936,
      "learning_rate": 0.00018403663229607652,
      "loss": 0.0537,
      "step": 1270
    },
    {
      "epoch": 0.8437705998681608,
      "grad_norm": 0.6518540382385254,
      "learning_rate": 0.00018371999591918723,
      "loss": 0.0679,
      "step": 1280
    },
    {
      "epoch": 0.8503625576796309,
      "grad_norm": 0.5521682500839233,
      "learning_rate": 0.0001834005280292963,
      "loss": 0.0609,
      "step": 1290
    },
    {
      "epoch": 0.8569545154911009,
      "grad_norm": 0.5179183483123779,
      "learning_rate": 0.00018307823943119976,
      "loss": 0.0623,
      "step": 1300
    },
    {
      "epoch": 0.8635464733025708,
      "grad_norm": 0.46778643131256104,
      "learning_rate": 0.00018275314102509356,
      "loss": 0.0543,
      "step": 1310
    },
    {
      "epoch": 0.8701384311140409,
      "grad_norm": 0.5073345303535461,
      "learning_rate": 0.00018242524380620472,
      "loss": 0.0498,
      "step": 1320
    },
    {
      "epoch": 0.8767303889255109,
      "grad_norm": 0.6397154331207275,
      "learning_rate": 0.00018209455886441958,
      "loss": 0.0638,
      "step": 1330
    },
    {
      "epoch": 0.8833223467369808,
      "grad_norm": 0.5524711012840271,
      "learning_rate": 0.00018176109738390865,
      "loss": 0.0697,
      "step": 1340
    },
    {
      "epoch": 0.8899143045484509,
      "grad_norm": 0.5283142328262329,
      "learning_rate": 0.00018142487064274848,
      "loss": 0.0535,
      "step": 1350
    },
    {
      "epoch": 0.8965062623599209,
      "grad_norm": 0.6182568669319153,
      "learning_rate": 0.00018108589001254,
      "loss": 0.075,
      "step": 1360
    },
    {
      "epoch": 0.9030982201713909,
      "grad_norm": 0.6030199527740479,
      "learning_rate": 0.0001807441669580242,
      "loss": 0.0507,
      "step": 1370
    },
    {
      "epoch": 0.9096901779828609,
      "grad_norm": 0.8080623149871826,
      "learning_rate": 0.00018039971303669407,
      "loss": 0.0573,
      "step": 1380
    },
    {
      "epoch": 0.9162821357943309,
      "grad_norm": 0.666598379611969,
      "learning_rate": 0.00018005253989840397,
      "loss": 0.0564,
      "step": 1390
    },
    {
      "epoch": 0.922874093605801,
      "grad_norm": 0.4886372983455658,
      "learning_rate": 0.00017970265928497542,
      "loss": 0.058,
      "step": 1400
    },
    {
      "epoch": 0.9294660514172709,
      "grad_norm": 0.5167548656463623,
      "learning_rate": 0.00017935008302980023,
      "loss": 0.0651,
      "step": 1410
    },
    {
      "epoch": 0.9360580092287409,
      "grad_norm": 0.48855462670326233,
      "learning_rate": 0.00017899482305743997,
      "loss": 0.0539,
      "step": 1420
    },
    {
      "epoch": 0.942649967040211,
      "grad_norm": 0.42956027388572693,
      "learning_rate": 0.00017863689138322282,
      "loss": 0.0517,
      "step": 1430
    },
    {
      "epoch": 0.9492419248516809,
      "grad_norm": 0.8002837896347046,
      "learning_rate": 0.00017827630011283739,
      "loss": 0.0417,
      "step": 1440
    },
    {
      "epoch": 0.955833882663151,
      "grad_norm": 0.6120493412017822,
      "learning_rate": 0.00017791306144192284,
      "loss": 0.0558,
      "step": 1450
    },
    {
      "epoch": 0.962425840474621,
      "grad_norm": 0.4339682161808014,
      "learning_rate": 0.00017754718765565681,
      "loss": 0.0502,
      "step": 1460
    },
    {
      "epoch": 0.9690177982860909,
      "grad_norm": 0.34096962213516235,
      "learning_rate": 0.00017717869112833978,
      "loss": 0.0362,
      "step": 1470
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.2626149356365204,
      "learning_rate": 0.00017680758432297644,
      "loss": 0.0505,
      "step": 1480
    },
    {
      "epoch": 0.982201713909031,
      "grad_norm": 0.46172085404396057,
      "learning_rate": 0.00017643387979085444,
      "loss": 0.0481,
      "step": 1490
    },
    {
      "epoch": 0.988793671720501,
      "grad_norm": 0.7232904434204102,
      "learning_rate": 0.00017605759017111963,
      "loss": 0.0609,
      "step": 1500
    },
    {
      "epoch": 0.995385629531971,
      "grad_norm": 0.594722330570221,
      "learning_rate": 0.00017567872819034875,
      "loss": 0.0457,
      "step": 1510
    },
    {
      "epoch": 1.0019775873434411,
      "grad_norm": 0.5150778889656067,
      "learning_rate": 0.00017529730666211886,
      "loss": 0.0508,
      "step": 1520
    },
    {
      "epoch": 1.008569545154911,
      "grad_norm": 0.7054618000984192,
      "learning_rate": 0.00017491333848657418,
      "loss": 0.0584,
      "step": 1530
    },
    {
      "epoch": 1.015161502966381,
      "grad_norm": 0.7400113940238953,
      "learning_rate": 0.00017452683664998954,
      "loss": 0.0479,
      "step": 1540
    },
    {
      "epoch": 1.0217534607778511,
      "grad_norm": 0.42146751284599304,
      "learning_rate": 0.00017413781422433145,
      "loss": 0.0593,
      "step": 1550
    },
    {
      "epoch": 1.028345418589321,
      "grad_norm": 0.43894773721694946,
      "learning_rate": 0.00017374628436681567,
      "loss": 0.0516,
      "step": 1560
    },
    {
      "epoch": 1.034937376400791,
      "grad_norm": 0.6406469941139221,
      "learning_rate": 0.0001733522603194625,
      "loss": 0.0426,
      "step": 1570
    },
    {
      "epoch": 1.0415293342122611,
      "grad_norm": 0.4033351540565491,
      "learning_rate": 0.00017295575540864877,
      "loss": 0.0508,
      "step": 1580
    },
    {
      "epoch": 1.048121292023731,
      "grad_norm": 0.4245661199092865,
      "learning_rate": 0.0001725567830446571,
      "loss": 0.0525,
      "step": 1590
    },
    {
      "epoch": 1.054713249835201,
      "grad_norm": 0.4368399381637573,
      "learning_rate": 0.0001721553567212225,
      "loss": 0.0525,
      "step": 1600
    },
    {
      "epoch": 1.0613052076466711,
      "grad_norm": 0.4312591552734375,
      "learning_rate": 0.00017175149001507581,
      "loss": 0.0501,
      "step": 1610
    },
    {
      "epoch": 1.067897165458141,
      "grad_norm": 0.43050646781921387,
      "learning_rate": 0.0001713451965854847,
      "loss": 0.0511,
      "step": 1620
    },
    {
      "epoch": 1.074489123269611,
      "grad_norm": 0.28264015913009644,
      "learning_rate": 0.0001709364901737915,
      "loss": 0.0407,
      "step": 1630
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 0.44433021545410156,
      "learning_rate": 0.00017052538460294865,
      "loss": 0.0434,
      "step": 1640
    },
    {
      "epoch": 1.087673038892551,
      "grad_norm": 0.40889009833335876,
      "learning_rate": 0.00017011189377705104,
      "loss": 0.0363,
      "step": 1650
    },
    {
      "epoch": 1.094264996704021,
      "grad_norm": 0.5670244693756104,
      "learning_rate": 0.00016969603168086583,
      "loss": 0.0526,
      "step": 1660
    },
    {
      "epoch": 1.1008569545154911,
      "grad_norm": 0.4179556965827942,
      "learning_rate": 0.00016927781237935946,
      "loss": 0.0444,
      "step": 1670
    },
    {
      "epoch": 1.107448912326961,
      "grad_norm": 0.49149006605148315,
      "learning_rate": 0.00016885725001722197,
      "loss": 0.0503,
      "step": 1680
    },
    {
      "epoch": 1.1140408701384312,
      "grad_norm": 0.5782753229141235,
      "learning_rate": 0.00016843435881838846,
      "loss": 0.0644,
      "step": 1690
    },
    {
      "epoch": 1.1206328279499012,
      "grad_norm": 0.3774556815624237,
      "learning_rate": 0.00016800915308555833,
      "loss": 0.0434,
      "step": 1700
    },
    {
      "epoch": 1.127224785761371,
      "grad_norm": 0.4292038679122925,
      "learning_rate": 0.00016758164719971123,
      "loss": 0.0369,
      "step": 1710
    },
    {
      "epoch": 1.133816743572841,
      "grad_norm": 0.4183276891708374,
      "learning_rate": 0.00016715185561962085,
      "loss": 0.0456,
      "step": 1720
    },
    {
      "epoch": 1.1404087013843112,
      "grad_norm": 0.5465393662452698,
      "learning_rate": 0.00016671979288136588,
      "loss": 0.0467,
      "step": 1730
    },
    {
      "epoch": 1.147000659195781,
      "grad_norm": 0.5326853394508362,
      "learning_rate": 0.0001662854735978383,
      "loss": 0.0541,
      "step": 1740
    },
    {
      "epoch": 1.1535926170072512,
      "grad_norm": 0.4676070809364319,
      "learning_rate": 0.00016584891245824933,
      "loss": 0.0409,
      "step": 1750
    },
    {
      "epoch": 1.1601845748187212,
      "grad_norm": 0.3939380645751953,
      "learning_rate": 0.00016541012422763242,
      "loss": 0.0405,
      "step": 1760
    },
    {
      "epoch": 1.166776532630191,
      "grad_norm": 0.4636197090148926,
      "learning_rate": 0.00016496912374634398,
      "loss": 0.047,
      "step": 1770
    },
    {
      "epoch": 1.1733684904416612,
      "grad_norm": 0.3863416612148285,
      "learning_rate": 0.0001645259259295615,
      "loss": 0.0411,
      "step": 1780
    },
    {
      "epoch": 1.1799604482531312,
      "grad_norm": 0.4277307093143463,
      "learning_rate": 0.00016408054576677906,
      "loss": 0.0436,
      "step": 1790
    },
    {
      "epoch": 1.186552406064601,
      "grad_norm": 0.3793676793575287,
      "learning_rate": 0.00016363299832130026,
      "loss": 0.0444,
      "step": 1800
    },
    {
      "epoch": 1.1931443638760713,
      "grad_norm": 0.4446391761302948,
      "learning_rate": 0.00016318329872972898,
      "loss": 0.0362,
      "step": 1810
    },
    {
      "epoch": 1.1997363216875412,
      "grad_norm": 0.44605711102485657,
      "learning_rate": 0.00016273146220145725,
      "loss": 0.032,
      "step": 1820
    },
    {
      "epoch": 1.2063282794990111,
      "grad_norm": 0.46649622917175293,
      "learning_rate": 0.00016227750401815092,
      "loss": 0.0367,
      "step": 1830
    },
    {
      "epoch": 1.2129202373104813,
      "grad_norm": 0.44259113073349,
      "learning_rate": 0.00016182143953323287,
      "loss": 0.0387,
      "step": 1840
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 0.519391655921936,
      "learning_rate": 0.0001613632841713637,
      "loss": 0.0384,
      "step": 1850
    },
    {
      "epoch": 1.2261041529334213,
      "grad_norm": 0.5217844843864441,
      "learning_rate": 0.00016090305342791989,
      "loss": 0.0366,
      "step": 1860
    },
    {
      "epoch": 1.2326961107448913,
      "grad_norm": 0.6881841421127319,
      "learning_rate": 0.00016044076286847,
      "loss": 0.0393,
      "step": 1870
    },
    {
      "epoch": 1.2392880685563612,
      "grad_norm": 0.39065277576446533,
      "learning_rate": 0.0001599764281282481,
      "loss": 0.0373,
      "step": 1880
    },
    {
      "epoch": 1.2458800263678311,
      "grad_norm": 0.5099205374717712,
      "learning_rate": 0.00015951006491162493,
      "loss": 0.0467,
      "step": 1890
    },
    {
      "epoch": 1.2524719841793013,
      "grad_norm": 0.6664186716079712,
      "learning_rate": 0.00015904168899157672,
      "loss": 0.0582,
      "step": 1900
    },
    {
      "epoch": 1.2590639419907712,
      "grad_norm": 0.5066046714782715,
      "learning_rate": 0.00015857131620915194,
      "loss": 0.0369,
      "step": 1910
    },
    {
      "epoch": 1.2656558998022414,
      "grad_norm": 0.3705768585205078,
      "learning_rate": 0.0001580989624729353,
      "loss": 0.0361,
      "step": 1920
    },
    {
      "epoch": 1.2722478576137113,
      "grad_norm": 0.5163792371749878,
      "learning_rate": 0.00015762464375850983,
      "loss": 0.0411,
      "step": 1930
    },
    {
      "epoch": 1.2788398154251812,
      "grad_norm": 0.4548538029193878,
      "learning_rate": 0.0001571483761079165,
      "loss": 0.0365,
      "step": 1940
    },
    {
      "epoch": 1.2854317732366514,
      "grad_norm": 0.3885802626609802,
      "learning_rate": 0.00015667017562911176,
      "loss": 0.0411,
      "step": 1950
    },
    {
      "epoch": 1.2920237310481213,
      "grad_norm": 0.4474758803844452,
      "learning_rate": 0.00015619005849542262,
      "loss": 0.0427,
      "step": 1960
    },
    {
      "epoch": 1.2986156888595912,
      "grad_norm": 0.33950087428092957,
      "learning_rate": 0.00015570804094499976,
      "loss": 0.0421,
      "step": 1970
    },
    {
      "epoch": 1.3052076466710614,
      "grad_norm": 0.33460333943367004,
      "learning_rate": 0.0001552241392802682,
      "loss": 0.0411,
      "step": 1980
    },
    {
      "epoch": 1.3117996044825313,
      "grad_norm": 0.5268707871437073,
      "learning_rate": 0.00015473836986737611,
      "loss": 0.0315,
      "step": 1990
    },
    {
      "epoch": 1.3183915622940012,
      "grad_norm": 0.42631861567497253,
      "learning_rate": 0.00015425074913564117,
      "loss": 0.0409,
      "step": 2000
    },
    {
      "epoch": 1.3249835201054714,
      "grad_norm": 0.3418896496295929,
      "learning_rate": 0.00015376129357699494,
      "loss": 0.0402,
      "step": 2010
    },
    {
      "epoch": 1.3315754779169413,
      "grad_norm": 0.43932780623435974,
      "learning_rate": 0.000153270019745425,
      "loss": 0.0408,
      "step": 2020
    },
    {
      "epoch": 1.3381674357284115,
      "grad_norm": 0.653154730796814,
      "learning_rate": 0.00015277694425641528,
      "loss": 0.0395,
      "step": 2030
    },
    {
      "epoch": 1.3447593935398814,
      "grad_norm": 0.49582821130752563,
      "learning_rate": 0.0001522820837863839,
      "loss": 0.0356,
      "step": 2040
    },
    {
      "epoch": 1.3513513513513513,
      "grad_norm": 0.24573160707950592,
      "learning_rate": 0.0001517854550721192,
      "loss": 0.0347,
      "step": 2050
    },
    {
      "epoch": 1.3579433091628212,
      "grad_norm": 0.2804608643054962,
      "learning_rate": 0.0001512870749102138,
      "loss": 0.04,
      "step": 2060
    },
    {
      "epoch": 1.3645352669742914,
      "grad_norm": 0.3183498680591583,
      "learning_rate": 0.00015078696015649639,
      "loss": 0.0388,
      "step": 2070
    },
    {
      "epoch": 1.3711272247857613,
      "grad_norm": 0.4484601318836212,
      "learning_rate": 0.0001502851277254616,
      "loss": 0.0376,
      "step": 2080
    },
    {
      "epoch": 1.3777191825972315,
      "grad_norm": 0.3878334164619446,
      "learning_rate": 0.00014978159458969815,
      "loss": 0.0349,
      "step": 2090
    },
    {
      "epoch": 1.3843111404087014,
      "grad_norm": 0.5986606478691101,
      "learning_rate": 0.0001492763777793146,
      "loss": 0.0333,
      "step": 2100
    },
    {
      "epoch": 1.3909030982201713,
      "grad_norm": 0.509537935256958,
      "learning_rate": 0.00014876949438136347,
      "loss": 0.0344,
      "step": 2110
    },
    {
      "epoch": 1.3974950560316415,
      "grad_norm": 0.5156903862953186,
      "learning_rate": 0.00014826096153926332,
      "loss": 0.0511,
      "step": 2120
    },
    {
      "epoch": 1.4040870138431114,
      "grad_norm": 0.5031445622444153,
      "learning_rate": 0.0001477507964522189,
      "loss": 0.0427,
      "step": 2130
    },
    {
      "epoch": 1.4106789716545813,
      "grad_norm": 0.603550910949707,
      "learning_rate": 0.00014723901637463956,
      "loss": 0.0379,
      "step": 2140
    },
    {
      "epoch": 1.4172709294660515,
      "grad_norm": 0.38619402050971985,
      "learning_rate": 0.0001467256386155556,
      "loss": 0.0365,
      "step": 2150
    },
    {
      "epoch": 1.4238628872775214,
      "grad_norm": 0.4471161365509033,
      "learning_rate": 0.00014621068053803272,
      "loss": 0.037,
      "step": 2160
    },
    {
      "epoch": 1.4304548450889913,
      "grad_norm": 0.5192623734474182,
      "learning_rate": 0.00014569415955858518,
      "loss": 0.0374,
      "step": 2170
    },
    {
      "epoch": 1.4370468029004615,
      "grad_norm": 0.3612404465675354,
      "learning_rate": 0.00014517609314658632,
      "loss": 0.0334,
      "step": 2180
    },
    {
      "epoch": 1.4436387607119314,
      "grad_norm": 0.5519164800643921,
      "learning_rate": 0.00014465649882367793,
      "loss": 0.036,
      "step": 2190
    },
    {
      "epoch": 1.4502307185234016,
      "grad_norm": 0.3743966221809387,
      "learning_rate": 0.00014413539416317767,
      "loss": 0.0316,
      "step": 2200
    },
    {
      "epoch": 1.4568226763348715,
      "grad_norm": 0.44635990262031555,
      "learning_rate": 0.00014361279678948467,
      "loss": 0.0419,
      "step": 2210
    },
    {
      "epoch": 1.4634146341463414,
      "grad_norm": 0.42854979634284973,
      "learning_rate": 0.0001430887243774834,
      "loss": 0.0332,
      "step": 2220
    },
    {
      "epoch": 1.4700065919578114,
      "grad_norm": 0.349597692489624,
      "learning_rate": 0.000142563194651946,
      "loss": 0.0379,
      "step": 2230
    },
    {
      "epoch": 1.4765985497692815,
      "grad_norm": 0.7277832627296448,
      "learning_rate": 0.00014203622538693262,
      "loss": 0.0357,
      "step": 2240
    },
    {
      "epoch": 1.4831905075807514,
      "grad_norm": 0.664358377456665,
      "learning_rate": 0.0001415078344051906,
      "loss": 0.0392,
      "step": 2250
    },
    {
      "epoch": 1.4897824653922216,
      "grad_norm": 0.7072487473487854,
      "learning_rate": 0.00014097803957755132,
      "loss": 0.0459,
      "step": 2260
    },
    {
      "epoch": 1.4963744232036915,
      "grad_norm": 0.426103413105011,
      "learning_rate": 0.000140446858822326,
      "loss": 0.0413,
      "step": 2270
    },
    {
      "epoch": 1.5029663810151614,
      "grad_norm": 0.9970699548721313,
      "learning_rate": 0.00013991431010469967,
      "loss": 0.0409,
      "step": 2280
    },
    {
      "epoch": 1.5095583388266314,
      "grad_norm": 0.4669429063796997,
      "learning_rate": 0.0001393804114361235,
      "loss": 0.0334,
      "step": 2290
    },
    {
      "epoch": 1.5161502966381015,
      "grad_norm": 0.8753983974456787,
      "learning_rate": 0.0001388451808737056,
      "loss": 0.0425,
      "step": 2300
    },
    {
      "epoch": 1.5227422544495717,
      "grad_norm": 0.3841276466846466,
      "learning_rate": 0.00013830863651960042,
      "loss": 0.0352,
      "step": 2310
    },
    {
      "epoch": 1.5293342122610416,
      "grad_norm": 0.6196696758270264,
      "learning_rate": 0.0001377707965203965,
      "loss": 0.0368,
      "step": 2320
    },
    {
      "epoch": 1.5359261700725115,
      "grad_norm": 0.3803444802761078,
      "learning_rate": 0.00013723167906650263,
      "loss": 0.0381,
      "step": 2330
    },
    {
      "epoch": 1.5425181278839815,
      "grad_norm": 0.4198368191719055,
      "learning_rate": 0.00013669130239153264,
      "loss": 0.0414,
      "step": 2340
    },
    {
      "epoch": 1.5491100856954514,
      "grad_norm": 0.5054610967636108,
      "learning_rate": 0.0001361496847716889,
      "loss": 0.0345,
      "step": 2350
    },
    {
      "epoch": 1.5557020435069215,
      "grad_norm": 0.46151942014694214,
      "learning_rate": 0.00013560684452514394,
      "loss": 0.0325,
      "step": 2360
    },
    {
      "epoch": 1.5622940013183917,
      "grad_norm": 0.5142484903335571,
      "learning_rate": 0.00013506280001142104,
      "loss": 0.034,
      "step": 2370
    },
    {
      "epoch": 1.5688859591298616,
      "grad_norm": 0.46030113101005554,
      "learning_rate": 0.0001345175696307733,
      "loss": 0.0271,
      "step": 2380
    },
    {
      "epoch": 1.5754779169413315,
      "grad_norm": 0.7596203088760376,
      "learning_rate": 0.00013397117182356136,
      "loss": 0.0389,
      "step": 2390
    },
    {
      "epoch": 1.5820698747528015,
      "grad_norm": 0.5491079688072205,
      "learning_rate": 0.00013342362506962952,
      "loss": 0.0348,
      "step": 2400
    },
    {
      "epoch": 1.5886618325642716,
      "grad_norm": 0.7162690162658691,
      "learning_rate": 0.00013287494788768095,
      "loss": 0.0377,
      "step": 2410
    },
    {
      "epoch": 1.5952537903757416,
      "grad_norm": 1.0456359386444092,
      "learning_rate": 0.00013232515883465125,
      "loss": 0.0364,
      "step": 2420
    },
    {
      "epoch": 1.6018457481872117,
      "grad_norm": 0.5690692067146301,
      "learning_rate": 0.00013177427650508093,
      "loss": 0.0338,
      "step": 2430
    },
    {
      "epoch": 1.6084377059986816,
      "grad_norm": 0.46750396490097046,
      "learning_rate": 0.0001312223195304864,
      "loss": 0.0357,
      "step": 2440
    },
    {
      "epoch": 1.6150296638101516,
      "grad_norm": 0.3841594159603119,
      "learning_rate": 0.00013066930657872986,
      "loss": 0.033,
      "step": 2450
    },
    {
      "epoch": 1.6216216216216215,
      "grad_norm": 0.46823808550834656,
      "learning_rate": 0.00013011525635338796,
      "loss": 0.0306,
      "step": 2460
    },
    {
      "epoch": 1.6282135794330916,
      "grad_norm": 0.5914480686187744,
      "learning_rate": 0.00012956018759311934,
      "loss": 0.0322,
      "step": 2470
    },
    {
      "epoch": 1.6348055372445618,
      "grad_norm": 0.2972632944583893,
      "learning_rate": 0.0001290041190710306,
      "loss": 0.0436,
      "step": 2480
    },
    {
      "epoch": 1.6413974950560317,
      "grad_norm": 0.3952536880970001,
      "learning_rate": 0.0001284470695940415,
      "loss": 0.0386,
      "step": 2490
    },
    {
      "epoch": 1.6479894528675016,
      "grad_norm": 0.42823290824890137,
      "learning_rate": 0.00012788905800224904,
      "loss": 0.0396,
      "step": 2500
    },
    {
      "epoch": 1.6545814106789716,
      "grad_norm": 0.39351433515548706,
      "learning_rate": 0.00012733010316828998,
      "loss": 0.0307,
      "step": 2510
    },
    {
      "epoch": 1.6611733684904415,
      "grad_norm": 0.23917658627033234,
      "learning_rate": 0.00012677022399670279,
      "loss": 0.0248,
      "step": 2520
    },
    {
      "epoch": 1.6677653263019117,
      "grad_norm": 0.3799072802066803,
      "learning_rate": 0.00012620943942328802,
      "loss": 0.0313,
      "step": 2530
    },
    {
      "epoch": 1.6743572841133818,
      "grad_norm": 0.36729612946510315,
      "learning_rate": 0.00012564776841446826,
      "loss": 0.0392,
      "step": 2540
    },
    {
      "epoch": 1.6809492419248517,
      "grad_norm": 0.45756933093070984,
      "learning_rate": 0.00012508522996664622,
      "loss": 0.0333,
      "step": 2550
    },
    {
      "epoch": 1.6875411997363217,
      "grad_norm": 0.6483779549598694,
      "learning_rate": 0.00012452184310556257,
      "loss": 0.0398,
      "step": 2560
    },
    {
      "epoch": 1.6941331575477916,
      "grad_norm": 0.3594001829624176,
      "learning_rate": 0.0001239576268856523,
      "loss": 0.0418,
      "step": 2570
    },
    {
      "epoch": 1.7007251153592617,
      "grad_norm": 0.7351908683776855,
      "learning_rate": 0.00012339260038940044,
      "loss": 0.0401,
      "step": 2580
    },
    {
      "epoch": 1.7073170731707317,
      "grad_norm": 0.6257510781288147,
      "learning_rate": 0.0001228267827266964,
      "loss": 0.0323,
      "step": 2590
    },
    {
      "epoch": 1.7139090309822018,
      "grad_norm": 0.2904309034347534,
      "learning_rate": 0.00012226019303418793,
      "loss": 0.0353,
      "step": 2600
    },
    {
      "epoch": 1.7205009887936717,
      "grad_norm": 0.403761625289917,
      "learning_rate": 0.00012169285047463372,
      "loss": 0.0303,
      "step": 2610
    },
    {
      "epoch": 1.7270929466051417,
      "grad_norm": 0.5491513013839722,
      "learning_rate": 0.00012112477423625539,
      "loss": 0.03,
      "step": 2620
    },
    {
      "epoch": 1.7336849044166116,
      "grad_norm": 0.6333141922950745,
      "learning_rate": 0.00012055598353208845,
      "loss": 0.0323,
      "step": 2630
    },
    {
      "epoch": 1.7402768622280818,
      "grad_norm": 0.5011347532272339,
      "learning_rate": 0.00011998649759933253,
      "loss": 0.0306,
      "step": 2640
    },
    {
      "epoch": 1.746868820039552,
      "grad_norm": 0.48753035068511963,
      "learning_rate": 0.00011941633569870073,
      "loss": 0.0251,
      "step": 2650
    },
    {
      "epoch": 1.7534607778510218,
      "grad_norm": 0.3967491388320923,
      "learning_rate": 0.0001188455171137682,
      "loss": 0.0285,
      "step": 2660
    },
    {
      "epoch": 1.7600527356624918,
      "grad_norm": 0.5975573062896729,
      "learning_rate": 0.00011827406115032005,
      "loss": 0.0355,
      "step": 2670
    },
    {
      "epoch": 1.7666446934739617,
      "grad_norm": 0.3887605369091034,
      "learning_rate": 0.00011770198713569814,
      "loss": 0.0293,
      "step": 2680
    },
    {
      "epoch": 1.7732366512854316,
      "grad_norm": 0.2963058650493622,
      "learning_rate": 0.00011712931441814776,
      "loss": 0.0229,
      "step": 2690
    },
    {
      "epoch": 1.7798286090969018,
      "grad_norm": 0.2195950299501419,
      "learning_rate": 0.00011655606236616302,
      "loss": 0.0285,
      "step": 2700
    },
    {
      "epoch": 1.786420566908372,
      "grad_norm": 0.3120424151420593,
      "learning_rate": 0.00011598225036783168,
      "loss": 0.0261,
      "step": 2710
    },
    {
      "epoch": 1.7930125247198418,
      "grad_norm": 0.23726621270179749,
      "learning_rate": 0.0001154078978301798,
      "loss": 0.03,
      "step": 2720
    },
    {
      "epoch": 1.7996044825313118,
      "grad_norm": 0.3860594928264618,
      "learning_rate": 0.00011483302417851503,
      "loss": 0.0279,
      "step": 2730
    },
    {
      "epoch": 1.8061964403427817,
      "grad_norm": 0.3965723514556885,
      "learning_rate": 0.00011425764885576968,
      "loss": 0.0291,
      "step": 2740
    },
    {
      "epoch": 1.8127883981542519,
      "grad_norm": 0.415278822183609,
      "learning_rate": 0.00011368179132184331,
      "loss": 0.0313,
      "step": 2750
    },
    {
      "epoch": 1.8193803559657218,
      "grad_norm": 0.43956270813941956,
      "learning_rate": 0.0001131054710529444,
      "loss": 0.0334,
      "step": 2760
    },
    {
      "epoch": 1.825972313777192,
      "grad_norm": 0.509265124797821,
      "learning_rate": 0.00011252870754093168,
      "loss": 0.0285,
      "step": 2770
    },
    {
      "epoch": 1.8325642715886619,
      "grad_norm": 0.373096227645874,
      "learning_rate": 0.00011195152029265492,
      "loss": 0.0323,
      "step": 2780
    },
    {
      "epoch": 1.8391562294001318,
      "grad_norm": 0.5025624632835388,
      "learning_rate": 0.00011137392882929515,
      "loss": 0.0266,
      "step": 2790
    },
    {
      "epoch": 1.8457481872116017,
      "grad_norm": 0.6312963962554932,
      "learning_rate": 0.00011079595268570451,
      "loss": 0.033,
      "step": 2800
    },
    {
      "epoch": 1.8523401450230719,
      "grad_norm": 0.525477409362793,
      "learning_rate": 0.00011021761140974547,
      "loss": 0.0323,
      "step": 2810
    },
    {
      "epoch": 1.858932102834542,
      "grad_norm": 0.5473213791847229,
      "learning_rate": 0.00010963892456162966,
      "loss": 0.0297,
      "step": 2820
    },
    {
      "epoch": 1.865524060646012,
      "grad_norm": 0.3462737202644348,
      "learning_rate": 0.00010905991171325645,
      "loss": 0.035,
      "step": 2830
    },
    {
      "epoch": 1.8721160184574819,
      "grad_norm": 0.41637200117111206,
      "learning_rate": 0.00010848059244755093,
      "loss": 0.0342,
      "step": 2840
    },
    {
      "epoch": 1.8787079762689518,
      "grad_norm": 0.4165598154067993,
      "learning_rate": 0.00010790098635780164,
      "loss": 0.0282,
      "step": 2850
    },
    {
      "epoch": 1.8852999340804217,
      "grad_norm": 0.5277098417282104,
      "learning_rate": 0.00010732111304699772,
      "loss": 0.0324,
      "step": 2860
    },
    {
      "epoch": 1.8918918918918919,
      "grad_norm": 0.6012358069419861,
      "learning_rate": 0.00010674099212716627,
      "loss": 0.0326,
      "step": 2870
    },
    {
      "epoch": 1.898483849703362,
      "grad_norm": 0.4933426082134247,
      "learning_rate": 0.00010616064321870862,
      "loss": 0.0261,
      "step": 2880
    },
    {
      "epoch": 1.905075807514832,
      "grad_norm": 0.5412283539772034,
      "learning_rate": 0.00010558008594973709,
      "loss": 0.0394,
      "step": 2890
    },
    {
      "epoch": 1.911667765326302,
      "grad_norm": 0.33404308557510376,
      "learning_rate": 0.00010499933995541096,
      "loss": 0.0324,
      "step": 2900
    },
    {
      "epoch": 1.9182597231377718,
      "grad_norm": 0.33016228675842285,
      "learning_rate": 0.00010441842487727247,
      "loss": 0.0291,
      "step": 2910
    },
    {
      "epoch": 1.924851680949242,
      "grad_norm": 0.3352147936820984,
      "learning_rate": 0.0001038373603625824,
      "loss": 0.0284,
      "step": 2920
    },
    {
      "epoch": 1.931443638760712,
      "grad_norm": 0.6947084665298462,
      "learning_rate": 0.00010325616606365571,
      "loss": 0.0338,
      "step": 2930
    },
    {
      "epoch": 1.938035596572182,
      "grad_norm": 0.5351916551589966,
      "learning_rate": 0.00010267486163719684,
      "loss": 0.037,
      "step": 2940
    },
    {
      "epoch": 1.944627554383652,
      "grad_norm": 0.3044882118701935,
      "learning_rate": 0.00010209346674363487,
      "loss": 0.0434,
      "step": 2950
    },
    {
      "epoch": 1.951219512195122,
      "grad_norm": 0.3456181287765503,
      "learning_rate": 0.00010151200104645855,
      "loss": 0.0313,
      "step": 2960
    },
    {
      "epoch": 1.9578114700065918,
      "grad_norm": 0.5266302824020386,
      "learning_rate": 0.00010093048421155136,
      "loss": 0.0315,
      "step": 2970
    },
    {
      "epoch": 1.964403427818062,
      "grad_norm": 0.43212220072746277,
      "learning_rate": 0.00010034893590652625,
      "loss": 0.0338,
      "step": 2980
    },
    {
      "epoch": 1.9709953856295321,
      "grad_norm": 0.48578178882598877,
      "learning_rate": 9.976737580006057e-05,
      "loss": 0.0351,
      "step": 2990
    },
    {
      "epoch": 1.977587343441002,
      "grad_norm": 0.38649991154670715,
      "learning_rate": 9.91858235612308e-05,
      "loss": 0.0376,
      "step": 3000
    },
    {
      "epoch": 1.984179301252472,
      "grad_norm": 0.392405241727829,
      "learning_rate": 9.860429885884736e-05,
      "loss": 0.0303,
      "step": 3010
    },
    {
      "epoch": 1.990771259063942,
      "grad_norm": 0.4340122640132904,
      "learning_rate": 9.802282136078927e-05,
      "loss": 0.031,
      "step": 3020
    },
    {
      "epoch": 1.9973632168754119,
      "grad_norm": 0.41526365280151367,
      "learning_rate": 9.74414107333391e-05,
      "loss": 0.036,
      "step": 3030
    },
    {
      "epoch": 2.0039551746868822,
      "grad_norm": 0.497365266084671,
      "learning_rate": 9.686008664051784e-05,
      "loss": 0.0329,
      "step": 3040
    },
    {
      "epoch": 2.010547132498352,
      "grad_norm": 0.29831039905548096,
      "learning_rate": 9.627886874341958e-05,
      "loss": 0.0387,
      "step": 3050
    },
    {
      "epoch": 2.017139090309822,
      "grad_norm": 0.5178902745246887,
      "learning_rate": 9.569777669954694e-05,
      "loss": 0.0333,
      "step": 3060
    },
    {
      "epoch": 2.023731048121292,
      "grad_norm": 0.42481425404548645,
      "learning_rate": 9.511683016214598e-05,
      "loss": 0.0285,
      "step": 3070
    },
    {
      "epoch": 2.030323005932762,
      "grad_norm": 0.38960328698158264,
      "learning_rate": 9.45360487795414e-05,
      "loss": 0.0304,
      "step": 3080
    },
    {
      "epoch": 2.036914963744232,
      "grad_norm": 0.2346969097852707,
      "learning_rate": 9.395545219447244e-05,
      "loss": 0.039,
      "step": 3090
    },
    {
      "epoch": 2.0435069215557022,
      "grad_norm": 0.3628898561000824,
      "learning_rate": 9.337506004342796e-05,
      "loss": 0.032,
      "step": 3100
    },
    {
      "epoch": 2.050098879367172,
      "grad_norm": 0.32154813408851624,
      "learning_rate": 9.279489195598281e-05,
      "loss": 0.0287,
      "step": 3110
    },
    {
      "epoch": 2.056690837178642,
      "grad_norm": 0.5084890723228455,
      "learning_rate": 9.221496755413375e-05,
      "loss": 0.0284,
      "step": 3120
    },
    {
      "epoch": 2.063282794990112,
      "grad_norm": 0.7246992588043213,
      "learning_rate": 9.163530645163562e-05,
      "loss": 0.0289,
      "step": 3130
    },
    {
      "epoch": 2.069874752801582,
      "grad_norm": 0.3673562705516815,
      "learning_rate": 9.105592825333829e-05,
      "loss": 0.025,
      "step": 3140
    },
    {
      "epoch": 2.076466710613052,
      "grad_norm": 0.25608980655670166,
      "learning_rate": 9.047685255452347e-05,
      "loss": 0.0302,
      "step": 3150
    },
    {
      "epoch": 2.0830586684245223,
      "grad_norm": 0.4780118763446808,
      "learning_rate": 8.989809894024186e-05,
      "loss": 0.0276,
      "step": 3160
    },
    {
      "epoch": 2.089650626235992,
      "grad_norm": 0.2582448124885559,
      "learning_rate": 8.931968698465096e-05,
      "loss": 0.0316,
      "step": 3170
    },
    {
      "epoch": 2.096242584047462,
      "grad_norm": 0.4123593270778656,
      "learning_rate": 8.874163625035295e-05,
      "loss": 0.0296,
      "step": 3180
    },
    {
      "epoch": 2.102834541858932,
      "grad_norm": 0.3350667357444763,
      "learning_rate": 8.8163966287733e-05,
      "loss": 0.024,
      "step": 3190
    },
    {
      "epoch": 2.109426499670402,
      "grad_norm": 0.518301784992218,
      "learning_rate": 8.758669663429818e-05,
      "loss": 0.0346,
      "step": 3200
    },
    {
      "epoch": 2.1160184574818723,
      "grad_norm": 0.25243598222732544,
      "learning_rate": 8.700984681401657e-05,
      "loss": 0.0266,
      "step": 3210
    },
    {
      "epoch": 2.1226104152933423,
      "grad_norm": 0.40096205472946167,
      "learning_rate": 8.643343633665697e-05,
      "loss": 0.0271,
      "step": 3220
    },
    {
      "epoch": 2.129202373104812,
      "grad_norm": 0.4391838014125824,
      "learning_rate": 8.585748469712913e-05,
      "loss": 0.0251,
      "step": 3230
    },
    {
      "epoch": 2.135794330916282,
      "grad_norm": 0.4758305251598358,
      "learning_rate": 8.528201137482426e-05,
      "loss": 0.0217,
      "step": 3240
    },
    {
      "epoch": 2.142386288727752,
      "grad_norm": 0.31953004002571106,
      "learning_rate": 8.470703583295634e-05,
      "loss": 0.0312,
      "step": 3250
    },
    {
      "epoch": 2.148978246539222,
      "grad_norm": 0.5065709352493286,
      "learning_rate": 8.413257751790382e-05,
      "loss": 0.0274,
      "step": 3260
    },
    {
      "epoch": 2.1555702043506924,
      "grad_norm": 0.44011905789375305,
      "learning_rate": 8.355865585855185e-05,
      "loss": 0.0284,
      "step": 3270
    },
    {
      "epoch": 2.1621621621621623,
      "grad_norm": 0.5650854706764221,
      "learning_rate": 8.298529026563525e-05,
      "loss": 0.0305,
      "step": 3280
    },
    {
      "epoch": 2.168754119973632,
      "grad_norm": 0.5405374765396118,
      "learning_rate": 8.241250013108204e-05,
      "loss": 0.0273,
      "step": 3290
    },
    {
      "epoch": 2.175346077785102,
      "grad_norm": 0.4947541654109955,
      "learning_rate": 8.184030482735744e-05,
      "loss": 0.0328,
      "step": 3300
    },
    {
      "epoch": 2.181938035596572,
      "grad_norm": 0.4064697027206421,
      "learning_rate": 8.126872370680889e-05,
      "loss": 0.0235,
      "step": 3310
    },
    {
      "epoch": 2.188529993408042,
      "grad_norm": 0.19350074231624603,
      "learning_rate": 8.069777610101117e-05,
      "loss": 0.0215,
      "step": 3320
    },
    {
      "epoch": 2.1951219512195124,
      "grad_norm": 0.22878113389015198,
      "learning_rate": 8.012748132011308e-05,
      "loss": 0.0226,
      "step": 3330
    },
    {
      "epoch": 2.2017139090309823,
      "grad_norm": 0.3672931492328644,
      "learning_rate": 7.955785865218399e-05,
      "loss": 0.027,
      "step": 3340
    },
    {
      "epoch": 2.2083058668424522,
      "grad_norm": 0.20331457257270813,
      "learning_rate": 7.898892736256147e-05,
      "loss": 0.0293,
      "step": 3350
    },
    {
      "epoch": 2.214897824653922,
      "grad_norm": 0.4426509439945221,
      "learning_rate": 7.842070669319994e-05,
      "loss": 0.0281,
      "step": 3360
    },
    {
      "epoch": 2.221489782465392,
      "grad_norm": 0.39120975136756897,
      "learning_rate": 7.785321586201983e-05,
      "loss": 0.0271,
      "step": 3370
    },
    {
      "epoch": 2.2280817402768625,
      "grad_norm": 0.3522750437259674,
      "learning_rate": 7.728647406225736e-05,
      "loss": 0.0257,
      "step": 3380
    },
    {
      "epoch": 2.2346736980883324,
      "grad_norm": 0.382185697555542,
      "learning_rate": 7.672050046181576e-05,
      "loss": 0.0227,
      "step": 3390
    },
    {
      "epoch": 2.2412656558998023,
      "grad_norm": 0.5848444104194641,
      "learning_rate": 7.615531420261669e-05,
      "loss": 0.0274,
      "step": 3400
    },
    {
      "epoch": 2.2478576137112722,
      "grad_norm": 0.39687180519104004,
      "learning_rate": 7.559093439995306e-05,
      "loss": 0.0306,
      "step": 3410
    },
    {
      "epoch": 2.254449571522742,
      "grad_norm": 0.4460280239582062,
      "learning_rate": 7.502738014184243e-05,
      "loss": 0.0319,
      "step": 3420
    },
    {
      "epoch": 2.261041529334212,
      "grad_norm": 0.23736679553985596,
      "learning_rate": 7.446467048838131e-05,
      "loss": 0.0223,
      "step": 3430
    },
    {
      "epoch": 2.267633487145682,
      "grad_norm": 0.4469570219516754,
      "learning_rate": 7.390282447110078e-05,
      "loss": 0.0367,
      "step": 3440
    },
    {
      "epoch": 2.2742254449571524,
      "grad_norm": 0.5411940813064575,
      "learning_rate": 7.334186109232264e-05,
      "loss": 0.0239,
      "step": 3450
    },
    {
      "epoch": 2.2808174027686223,
      "grad_norm": 0.5765545964241028,
      "learning_rate": 7.278179932451673e-05,
      "loss": 0.0253,
      "step": 3460
    },
    {
      "epoch": 2.2874093605800923,
      "grad_norm": 0.541907548904419,
      "learning_rate": 7.222265810965935e-05,
      "loss": 0.0331,
      "step": 3470
    },
    {
      "epoch": 2.294001318391562,
      "grad_norm": 0.2490926831960678,
      "learning_rate": 7.166445635859257e-05,
      "loss": 0.0222,
      "step": 3480
    },
    {
      "epoch": 2.300593276203032,
      "grad_norm": 0.3459300100803375,
      "learning_rate": 7.110721295038459e-05,
      "loss": 0.031,
      "step": 3490
    },
    {
      "epoch": 2.3071852340145025,
      "grad_norm": 0.2858792543411255,
      "learning_rate": 7.055094673169137e-05,
      "loss": 0.0227,
      "step": 3500
    },
    {
      "epoch": 2.3137771918259724,
      "grad_norm": 0.23397736251354218,
      "learning_rate": 6.9995676516119e-05,
      "loss": 0.0324,
      "step": 3510
    },
    {
      "epoch": 2.3203691496374423,
      "grad_norm": 0.41427797079086304,
      "learning_rate": 6.944142108358764e-05,
      "loss": 0.0274,
      "step": 3520
    },
    {
      "epoch": 2.3269611074489123,
      "grad_norm": 0.3530319631099701,
      "learning_rate": 6.888819917969618e-05,
      "loss": 0.0259,
      "step": 3530
    },
    {
      "epoch": 2.333553065260382,
      "grad_norm": 0.6433699727058411,
      "learning_rate": 6.833602951508827e-05,
      "loss": 0.0264,
      "step": 3540
    },
    {
      "epoch": 2.3401450230718526,
      "grad_norm": 0.41642090678215027,
      "learning_rate": 6.778493076481964e-05,
      "loss": 0.0319,
      "step": 3550
    },
    {
      "epoch": 2.3467369808833225,
      "grad_norm": 0.5816253423690796,
      "learning_rate": 6.723492156772633e-05,
      "loss": 0.0282,
      "step": 3560
    },
    {
      "epoch": 2.3533289386947924,
      "grad_norm": 0.4890478551387787,
      "learning_rate": 6.668602052579424e-05,
      "loss": 0.032,
      "step": 3570
    },
    {
      "epoch": 2.3599208965062624,
      "grad_norm": 0.4867573380470276,
      "learning_rate": 6.613824620353032e-05,
      "loss": 0.0284,
      "step": 3580
    },
    {
      "epoch": 2.3665128543177323,
      "grad_norm": 0.2637820541858673,
      "learning_rate": 6.559161712733437e-05,
      "loss": 0.0258,
      "step": 3590
    },
    {
      "epoch": 2.373104812129202,
      "grad_norm": 0.2814655601978302,
      "learning_rate": 6.504615178487246e-05,
      "loss": 0.0288,
      "step": 3600
    },
    {
      "epoch": 2.379696769940672,
      "grad_norm": 0.3690190315246582,
      "learning_rate": 6.450186862445197e-05,
      "loss": 0.0296,
      "step": 3610
    },
    {
      "epoch": 2.3862887277521425,
      "grad_norm": 0.47703316807746887,
      "learning_rate": 6.395878605439725e-05,
      "loss": 0.0284,
      "step": 3620
    },
    {
      "epoch": 2.3928806855636124,
      "grad_norm": 0.42962104082107544,
      "learning_rate": 6.341692244242727e-05,
      "loss": 0.0279,
      "step": 3630
    },
    {
      "epoch": 2.3994726433750824,
      "grad_norm": 0.38502004742622375,
      "learning_rate": 6.287629611503446e-05,
      "loss": 0.02,
      "step": 3640
    },
    {
      "epoch": 2.4060646011865523,
      "grad_norm": 0.39698663353919983,
      "learning_rate": 6.23369253568645e-05,
      "loss": 0.026,
      "step": 3650
    },
    {
      "epoch": 2.4126565589980222,
      "grad_norm": 0.34033870697021484,
      "learning_rate": 6.179882841009845e-05,
      "loss": 0.0226,
      "step": 3660
    },
    {
      "epoch": 2.4192485168094926,
      "grad_norm": 0.2257242500782013,
      "learning_rate": 6.126202347383537e-05,
      "loss": 0.0249,
      "step": 3670
    },
    {
      "epoch": 2.4258404746209625,
      "grad_norm": 0.44853436946868896,
      "learning_rate": 6.0726528703476935e-05,
      "loss": 0.0268,
      "step": 3680
    },
    {
      "epoch": 2.4324324324324325,
      "grad_norm": 0.3105967044830322,
      "learning_rate": 6.0192362210113454e-05,
      "loss": 0.0268,
      "step": 3690
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 0.19689629971981049,
      "learning_rate": 5.9659542059911266e-05,
      "loss": 0.0241,
      "step": 3700
    },
    {
      "epoch": 2.4456163480553723,
      "grad_norm": 0.3638206422328949,
      "learning_rate": 5.912808627350166e-05,
      "loss": 0.0286,
      "step": 3710
    },
    {
      "epoch": 2.4522083058668427,
      "grad_norm": 0.24674852192401886,
      "learning_rate": 5.8598012825371594e-05,
      "loss": 0.0279,
      "step": 3720
    },
    {
      "epoch": 2.4588002636783126,
      "grad_norm": 0.46596574783325195,
      "learning_rate": 5.8069339643255526e-05,
      "loss": 0.0277,
      "step": 3730
    },
    {
      "epoch": 2.4653922214897825,
      "grad_norm": 0.3376254737377167,
      "learning_rate": 5.7542084607529254e-05,
      "loss": 0.026,
      "step": 3740
    },
    {
      "epoch": 2.4719841793012525,
      "grad_norm": 0.3310098946094513,
      "learning_rate": 5.701626555060523e-05,
      "loss": 0.0237,
      "step": 3750
    },
    {
      "epoch": 2.4785761371127224,
      "grad_norm": 0.276920348405838,
      "learning_rate": 5.6491900256329063e-05,
      "loss": 0.0216,
      "step": 3760
    },
    {
      "epoch": 2.4851680949241923,
      "grad_norm": 0.26669642329216003,
      "learning_rate": 5.596900645937859e-05,
      "loss": 0.0218,
      "step": 3770
    },
    {
      "epoch": 2.4917600527356623,
      "grad_norm": 0.34589439630508423,
      "learning_rate": 5.5447601844663786e-05,
      "loss": 0.0231,
      "step": 3780
    },
    {
      "epoch": 2.4983520105471326,
      "grad_norm": 0.36905866861343384,
      "learning_rate": 5.4927704046728426e-05,
      "loss": 0.0243,
      "step": 3790
    },
    {
      "epoch": 2.5049439683586026,
      "grad_norm": 0.36546382308006287,
      "learning_rate": 5.440933064915414e-05,
      "loss": 0.0285,
      "step": 3800
    },
    {
      "epoch": 2.5115359261700725,
      "grad_norm": 0.6371796727180481,
      "learning_rate": 5.389249918396535e-05,
      "loss": 0.0293,
      "step": 3810
    },
    {
      "epoch": 2.5181278839815424,
      "grad_norm": 0.3491004705429077,
      "learning_rate": 5.3377227131036426e-05,
      "loss": 0.0262,
      "step": 3820
    },
    {
      "epoch": 2.5247198417930123,
      "grad_norm": 0.5807764530181885,
      "learning_rate": 5.28635319175006e-05,
      "loss": 0.0241,
      "step": 3830
    },
    {
      "epoch": 2.5313117996044827,
      "grad_norm": 0.33311110734939575,
      "learning_rate": 5.2351430917160327e-05,
      "loss": 0.0298,
      "step": 3840
    },
    {
      "epoch": 2.5379037574159526,
      "grad_norm": 0.42765146493911743,
      "learning_rate": 5.184094144989988e-05,
      "loss": 0.0215,
      "step": 3850
    },
    {
      "epoch": 2.5444957152274226,
      "grad_norm": 0.4080013036727905,
      "learning_rate": 5.1332080781099565e-05,
      "loss": 0.0267,
      "step": 3860
    },
    {
      "epoch": 2.5510876730388925,
      "grad_norm": 0.3769298195838928,
      "learning_rate": 5.082486612105164e-05,
      "loss": 0.025,
      "step": 3870
    },
    {
      "epoch": 2.5576796308503624,
      "grad_norm": 0.36813291907310486,
      "learning_rate": 5.031931462437829e-05,
      "loss": 0.0218,
      "step": 3880
    },
    {
      "epoch": 2.564271588661833,
      "grad_norm": 0.5158575177192688,
      "learning_rate": 4.981544338945163e-05,
      "loss": 0.0235,
      "step": 3890
    },
    {
      "epoch": 2.5708635464733027,
      "grad_norm": 0.26838409900665283,
      "learning_rate": 4.9313269457815124e-05,
      "loss": 0.0242,
      "step": 3900
    },
    {
      "epoch": 2.5774555042847727,
      "grad_norm": 0.3350141644477844,
      "learning_rate": 4.8812809813607366e-05,
      "loss": 0.0223,
      "step": 3910
    },
    {
      "epoch": 2.5840474620962426,
      "grad_norm": 0.43595895171165466,
      "learning_rate": 4.831408138298774e-05,
      "loss": 0.0178,
      "step": 3920
    },
    {
      "epoch": 2.5906394199077125,
      "grad_norm": 0.5088134407997131,
      "learning_rate": 4.7817101033563785e-05,
      "loss": 0.0257,
      "step": 3930
    },
    {
      "epoch": 2.5972313777191824,
      "grad_norm": 0.3459358811378479,
      "learning_rate": 4.732188557382078e-05,
      "loss": 0.0326,
      "step": 3940
    },
    {
      "epoch": 2.6038233355306524,
      "grad_norm": 0.5011377334594727,
      "learning_rate": 4.682845175255326e-05,
      "loss": 0.0263,
      "step": 3950
    },
    {
      "epoch": 2.6104152933421227,
      "grad_norm": 0.3368130624294281,
      "learning_rate": 4.633681625829869e-05,
      "loss": 0.0274,
      "step": 3960
    },
    {
      "epoch": 2.6170072511535927,
      "grad_norm": 0.3485671281814575,
      "learning_rate": 4.584699571877275e-05,
      "loss": 0.0237,
      "step": 3970
    },
    {
      "epoch": 2.6235992089650626,
      "grad_norm": 0.22069136798381805,
      "learning_rate": 4.535900670030715e-05,
      "loss": 0.0199,
      "step": 3980
    },
    {
      "epoch": 2.6301911667765325,
      "grad_norm": 0.21239358186721802,
      "learning_rate": 4.487286570728944e-05,
      "loss": 0.0198,
      "step": 3990
    },
    {
      "epoch": 2.6367831245880025,
      "grad_norm": 0.3069987893104553,
      "learning_rate": 4.438858918160452e-05,
      "loss": 0.0236,
      "step": 4000
    },
    {
      "epoch": 2.643375082399473,
      "grad_norm": 0.21374328434467316,
      "learning_rate": 4.390619350207882e-05,
      "loss": 0.0231,
      "step": 4010
    },
    {
      "epoch": 2.6499670402109428,
      "grad_norm": 0.2509225010871887,
      "learning_rate": 4.342569498392615e-05,
      "loss": 0.0192,
      "step": 4020
    },
    {
      "epoch": 2.6565589980224127,
      "grad_norm": 0.3323153257369995,
      "learning_rate": 4.294710987819612e-05,
      "loss": 0.028,
      "step": 4030
    },
    {
      "epoch": 2.6631509558338826,
      "grad_norm": 0.28733029961586,
      "learning_rate": 4.24704543712243e-05,
      "loss": 0.0261,
      "step": 4040
    },
    {
      "epoch": 2.6697429136453525,
      "grad_norm": 0.4628933370113373,
      "learning_rate": 4.199574458408487e-05,
      "loss": 0.0275,
      "step": 4050
    },
    {
      "epoch": 2.676334871456823,
      "grad_norm": 0.27761244773864746,
      "learning_rate": 4.1522996572045345e-05,
      "loss": 0.0278,
      "step": 4060
    },
    {
      "epoch": 2.682926829268293,
      "grad_norm": 0.31634223461151123,
      "learning_rate": 4.1052226324023724e-05,
      "loss": 0.0243,
      "step": 4070
    },
    {
      "epoch": 2.6895187870797628,
      "grad_norm": 0.2185506969690323,
      "learning_rate": 4.0583449762047464e-05,
      "loss": 0.0252,
      "step": 4080
    },
    {
      "epoch": 2.6961107448912327,
      "grad_norm": 0.43438541889190674,
      "learning_rate": 4.011668274071514e-05,
      "loss": 0.0213,
      "step": 4090
    },
    {
      "epoch": 2.7027027027027026,
      "grad_norm": 0.33553093671798706,
      "learning_rate": 3.965194104666029e-05,
      "loss": 0.0207,
      "step": 4100
    },
    {
      "epoch": 2.7092946605141726,
      "grad_norm": 0.21318501234054565,
      "learning_rate": 3.91892403980173e-05,
      "loss": 0.0225,
      "step": 4110
    },
    {
      "epoch": 2.7158866183256425,
      "grad_norm": 0.2645362615585327,
      "learning_rate": 3.872859644388989e-05,
      "loss": 0.0219,
      "step": 4120
    },
    {
      "epoch": 2.722478576137113,
      "grad_norm": 0.21204468607902527,
      "learning_rate": 3.827002476382193e-05,
      "loss": 0.0222,
      "step": 4130
    },
    {
      "epoch": 2.729070533948583,
      "grad_norm": 0.20982545614242554,
      "learning_rate": 3.781354086727038e-05,
      "loss": 0.0233,
      "step": 4140
    },
    {
      "epoch": 2.7356624917600527,
      "grad_norm": 0.41608384251594543,
      "learning_rate": 3.735916019308078e-05,
      "loss": 0.0288,
      "step": 4150
    },
    {
      "epoch": 2.7422544495715226,
      "grad_norm": 0.5119547843933105,
      "learning_rate": 3.690689810896518e-05,
      "loss": 0.0231,
      "step": 4160
    },
    {
      "epoch": 2.7488464073829926,
      "grad_norm": 0.3134307861328125,
      "learning_rate": 3.645676991098227e-05,
      "loss": 0.0216,
      "step": 4170
    },
    {
      "epoch": 2.755438365194463,
      "grad_norm": 0.26220589876174927,
      "learning_rate": 3.6008790823020043e-05,
      "loss": 0.022,
      "step": 4180
    },
    {
      "epoch": 2.762030323005933,
      "grad_norm": 0.41492775082588196,
      "learning_rate": 3.5562975996281064e-05,
      "loss": 0.0201,
      "step": 4190
    },
    {
      "epoch": 2.768622280817403,
      "grad_norm": 0.29042404890060425,
      "learning_rate": 3.511934050876986e-05,
      "loss": 0.0172,
      "step": 4200
    },
    {
      "epoch": 2.7752142386288727,
      "grad_norm": 0.20791085064411163,
      "learning_rate": 3.4677899364783e-05,
      "loss": 0.0232,
      "step": 4210
    },
    {
      "epoch": 2.7818061964403427,
      "grad_norm": 0.218453049659729,
      "learning_rate": 3.4238667494401786e-05,
      "loss": 0.0188,
      "step": 4220
    },
    {
      "epoch": 2.788398154251813,
      "grad_norm": 0.3837571144104004,
      "learning_rate": 3.380165975298697e-05,
      "loss": 0.0229,
      "step": 4230
    },
    {
      "epoch": 2.794990112063283,
      "grad_norm": 0.39384233951568604,
      "learning_rate": 3.336689092067673e-05,
      "loss": 0.0195,
      "step": 4240
    },
    {
      "epoch": 2.801582069874753,
      "grad_norm": 0.42460018396377563,
      "learning_rate": 3.2934375701886566e-05,
      "loss": 0.0306,
      "step": 4250
    },
    {
      "epoch": 2.808174027686223,
      "grad_norm": 0.21645621955394745,
      "learning_rate": 3.2504128724811835e-05,
      "loss": 0.0184,
      "step": 4260
    },
    {
      "epoch": 2.8147659854976927,
      "grad_norm": 0.3433361053466797,
      "learning_rate": 3.207616454093337e-05,
      "loss": 0.0177,
      "step": 4270
    },
    {
      "epoch": 2.8213579433091627,
      "grad_norm": 0.4108859598636627,
      "learning_rate": 3.1650497624525046e-05,
      "loss": 0.0246,
      "step": 4280
    },
    {
      "epoch": 2.8279499011206326,
      "grad_norm": 0.45586344599723816,
      "learning_rate": 3.122714237216431e-05,
      "loss": 0.0325,
      "step": 4290
    },
    {
      "epoch": 2.834541858932103,
      "grad_norm": 0.2443486601114273,
      "learning_rate": 3.080611310224539e-05,
      "loss": 0.0191,
      "step": 4300
    },
    {
      "epoch": 2.841133816743573,
      "grad_norm": 0.3153115510940552,
      "learning_rate": 3.038742405449485e-05,
      "loss": 0.0206,
      "step": 4310
    },
    {
      "epoch": 2.847725774555043,
      "grad_norm": 0.2398747354745865,
      "learning_rate": 2.997108938949006e-05,
      "loss": 0.0224,
      "step": 4320
    },
    {
      "epoch": 2.8543177323665128,
      "grad_norm": 0.3573615252971649,
      "learning_rate": 2.9557123188180358e-05,
      "loss": 0.0245,
      "step": 4330
    },
    {
      "epoch": 2.8609096901779827,
      "grad_norm": 0.5803711414337158,
      "learning_rate": 2.9145539451410675e-05,
      "loss": 0.0197,
      "step": 4340
    },
    {
      "epoch": 2.867501647989453,
      "grad_norm": 0.3499305248260498,
      "learning_rate": 2.8736352099448028e-05,
      "loss": 0.019,
      "step": 4350
    },
    {
      "epoch": 2.874093605800923,
      "grad_norm": 0.31976768374443054,
      "learning_rate": 2.8329574971510886e-05,
      "loss": 0.0219,
      "step": 4360
    },
    {
      "epoch": 2.880685563612393,
      "grad_norm": 0.44260144233703613,
      "learning_rate": 2.7925221825300852e-05,
      "loss": 0.0229,
      "step": 4370
    },
    {
      "epoch": 2.887277521423863,
      "grad_norm": 0.31007763743400574,
      "learning_rate": 2.7523306336537568e-05,
      "loss": 0.0174,
      "step": 4380
    },
    {
      "epoch": 2.8938694792353328,
      "grad_norm": 0.3109738826751709,
      "learning_rate": 2.712384209849603e-05,
      "loss": 0.0224,
      "step": 4390
    },
    {
      "epoch": 2.900461437046803,
      "grad_norm": 0.2678034007549286,
      "learning_rate": 2.672684262154709e-05,
      "loss": 0.0218,
      "step": 4400
    },
    {
      "epoch": 2.9070533948582726,
      "grad_norm": 0.31142377853393555,
      "learning_rate": 2.6332321332700172e-05,
      "loss": 0.0145,
      "step": 4410
    },
    {
      "epoch": 2.913645352669743,
      "grad_norm": 0.19287629425525665,
      "learning_rate": 2.59402915751494e-05,
      "loss": 0.0174,
      "step": 4420
    },
    {
      "epoch": 2.920237310481213,
      "grad_norm": 0.5407594442367554,
      "learning_rate": 2.5550766607822342e-05,
      "loss": 0.0201,
      "step": 4430
    },
    {
      "epoch": 2.926829268292683,
      "grad_norm": 0.2303568720817566,
      "learning_rate": 2.516375960493136e-05,
      "loss": 0.0194,
      "step": 4440
    },
    {
      "epoch": 2.933421226104153,
      "grad_norm": 0.2476644515991211,
      "learning_rate": 2.4779283655528195e-05,
      "loss": 0.0206,
      "step": 4450
    },
    {
      "epoch": 2.9400131839156227,
      "grad_norm": 0.2864633798599243,
      "learning_rate": 2.439735176306135e-05,
      "loss": 0.0217,
      "step": 4460
    },
    {
      "epoch": 2.946605141727093,
      "grad_norm": 0.5562433004379272,
      "learning_rate": 2.4017976844936084e-05,
      "loss": 0.0226,
      "step": 4470
    },
    {
      "epoch": 2.953197099538563,
      "grad_norm": 0.3683149516582489,
      "learning_rate": 2.364117173207766e-05,
      "loss": 0.021,
      "step": 4480
    },
    {
      "epoch": 2.959789057350033,
      "grad_norm": 0.5093694925308228,
      "learning_rate": 2.326694916849751e-05,
      "loss": 0.0194,
      "step": 4490
    },
    {
      "epoch": 2.966381015161503,
      "grad_norm": 0.16352635622024536,
      "learning_rate": 2.2895321810861837e-05,
      "loss": 0.0255,
      "step": 4500
    },
    {
      "epoch": 2.972972972972973,
      "grad_norm": 0.32750794291496277,
      "learning_rate": 2.252630222806399e-05,
      "loss": 0.0162,
      "step": 4510
    },
    {
      "epoch": 2.979564930784443,
      "grad_norm": 0.9548207521438599,
      "learning_rate": 2.215990290079918e-05,
      "loss": 0.0262,
      "step": 4520
    },
    {
      "epoch": 2.986156888595913,
      "grad_norm": 0.33009737730026245,
      "learning_rate": 2.1796136221142184e-05,
      "loss": 0.0201,
      "step": 4530
    },
    {
      "epoch": 2.992748846407383,
      "grad_norm": 0.30861908197402954,
      "learning_rate": 2.1435014492128547e-05,
      "loss": 0.0186,
      "step": 4540
    },
    {
      "epoch": 2.999340804218853,
      "grad_norm": 0.37097203731536865,
      "learning_rate": 2.1076549927338397e-05,
      "loss": 0.0233,
      "step": 4550
    },
    {
      "epoch": 3.005932762030323,
      "grad_norm": 0.29048582911491394,
      "learning_rate": 2.0720754650483076e-05,
      "loss": 0.0229,
      "step": 4560
    },
    {
      "epoch": 3.012524719841793,
      "grad_norm": 0.3153735399246216,
      "learning_rate": 2.036764069499555e-05,
      "loss": 0.0229,
      "step": 4570
    },
    {
      "epoch": 3.019116677653263,
      "grad_norm": 0.3723648488521576,
      "learning_rate": 2.001722000362317e-05,
      "loss": 0.025,
      "step": 4580
    },
    {
      "epoch": 3.025708635464733,
      "grad_norm": 0.3493257164955139,
      "learning_rate": 1.9669504428023644e-05,
      "loss": 0.0161,
      "step": 4590
    },
    {
      "epoch": 3.032300593276203,
      "grad_norm": 0.4804138243198395,
      "learning_rate": 1.9324505728364527e-05,
      "loss": 0.0199,
      "step": 4600
    },
    {
      "epoch": 3.038892551087673,
      "grad_norm": 0.40249496698379517,
      "learning_rate": 1.898223557292519e-05,
      "loss": 0.0239,
      "step": 4610
    },
    {
      "epoch": 3.045484508899143,
      "grad_norm": 0.4595593810081482,
      "learning_rate": 1.86427055377023e-05,
      "loss": 0.0255,
      "step": 4620
    },
    {
      "epoch": 3.052076466710613,
      "grad_norm": 0.22908352315425873,
      "learning_rate": 1.830592710601835e-05,
      "loss": 0.0165,
      "step": 4630
    },
    {
      "epoch": 3.058668424522083,
      "grad_norm": 0.2505341172218323,
      "learning_rate": 1.797191166813319e-05,
      "loss": 0.0186,
      "step": 4640
    },
    {
      "epoch": 3.065260382333553,
      "grad_norm": 0.20365647971630096,
      "learning_rate": 1.764067052085877e-05,
      "loss": 0.021,
      "step": 4650
    },
    {
      "epoch": 3.071852340145023,
      "grad_norm": 0.2220371514558792,
      "learning_rate": 1.7312214867177246e-05,
      "loss": 0.0252,
      "step": 4660
    },
    {
      "epoch": 3.078444297956493,
      "grad_norm": 0.2889988124370575,
      "learning_rate": 1.6986555815861882e-05,
      "loss": 0.0318,
      "step": 4670
    },
    {
      "epoch": 3.085036255767963,
      "grad_norm": 0.4432000517845154,
      "learning_rate": 1.666370438110141e-05,
      "loss": 0.0163,
      "step": 4680
    },
    {
      "epoch": 3.0916282135794333,
      "grad_norm": 0.44843411445617676,
      "learning_rate": 1.6343671482127597e-05,
      "loss": 0.0265,
      "step": 4690
    },
    {
      "epoch": 3.098220171390903,
      "grad_norm": 0.2467808723449707,
      "learning_rate": 1.6026467942845804e-05,
      "loss": 0.019,
      "step": 4700
    },
    {
      "epoch": 3.104812129202373,
      "grad_norm": 0.26013877987861633,
      "learning_rate": 1.5712104491468983e-05,
      "loss": 0.024,
      "step": 4710
    },
    {
      "epoch": 3.111404087013843,
      "grad_norm": 0.32995596528053284,
      "learning_rate": 1.5400591760154826e-05,
      "loss": 0.0202,
      "step": 4720
    },
    {
      "epoch": 3.117996044825313,
      "grad_norm": 0.3786914050579071,
      "learning_rate": 1.5091940284646245e-05,
      "loss": 0.0183,
      "step": 4730
    },
    {
      "epoch": 3.124588002636783,
      "grad_norm": 0.46880772709846497,
      "learning_rate": 1.478616050391488e-05,
      "loss": 0.0222,
      "step": 4740
    },
    {
      "epoch": 3.1311799604482533,
      "grad_norm": 0.24594926834106445,
      "learning_rate": 1.4483262759808136e-05,
      "loss": 0.0216,
      "step": 4750
    },
    {
      "epoch": 3.1377719182597232,
      "grad_norm": 0.40315452218055725,
      "learning_rate": 1.4183257296699493e-05,
      "loss": 0.0171,
      "step": 4760
    },
    {
      "epoch": 3.144363876071193,
      "grad_norm": 0.4592204988002777,
      "learning_rate": 1.3886154261141826e-05,
      "loss": 0.0186,
      "step": 4770
    },
    {
      "epoch": 3.150955833882663,
      "grad_norm": 0.413077175617218,
      "learning_rate": 1.3591963701524401e-05,
      "loss": 0.0296,
      "step": 4780
    },
    {
      "epoch": 3.157547791694133,
      "grad_norm": 0.2753660976886749,
      "learning_rate": 1.330069556773299e-05,
      "loss": 0.0258,
      "step": 4790
    },
    {
      "epoch": 3.164139749505603,
      "grad_norm": 0.33898794651031494,
      "learning_rate": 1.3012359710813304e-05,
      "loss": 0.0199,
      "step": 4800
    },
    {
      "epoch": 3.1707317073170733,
      "grad_norm": 0.2542150914669037,
      "learning_rate": 1.2726965882637853e-05,
      "loss": 0.0187,
      "step": 4810
    },
    {
      "epoch": 3.1773236651285433,
      "grad_norm": 0.2193000614643097,
      "learning_rate": 1.2444523735576197e-05,
      "loss": 0.0236,
      "step": 4820
    },
    {
      "epoch": 3.183915622940013,
      "grad_norm": 0.20990586280822754,
      "learning_rate": 1.216504282216826e-05,
      "loss": 0.017,
      "step": 4830
    },
    {
      "epoch": 3.190507580751483,
      "grad_norm": 0.30340689420700073,
      "learning_rate": 1.1888532594801583e-05,
      "loss": 0.0175,
      "step": 4840
    },
    {
      "epoch": 3.197099538562953,
      "grad_norm": 0.21258412301540375,
      "learning_rate": 1.1615002405391351e-05,
      "loss": 0.0182,
      "step": 4850
    },
    {
      "epoch": 3.2036914963744234,
      "grad_norm": 0.21666236221790314,
      "learning_rate": 1.134446150506423e-05,
      "loss": 0.0201,
      "step": 4860
    },
    {
      "epoch": 3.2102834541858933,
      "grad_norm": 0.44433897733688354,
      "learning_rate": 1.1076919043845513e-05,
      "loss": 0.0228,
      "step": 4870
    },
    {
      "epoch": 3.2168754119973633,
      "grad_norm": 0.35092124342918396,
      "learning_rate": 1.0812384070349535e-05,
      "loss": 0.02,
      "step": 4880
    },
    {
      "epoch": 3.223467369808833,
      "grad_norm": 0.33713802695274353,
      "learning_rate": 1.055086553147373e-05,
      "loss": 0.0193,
      "step": 4890
    },
    {
      "epoch": 3.230059327620303,
      "grad_norm": 0.45195287466049194,
      "learning_rate": 1.0292372272096029e-05,
      "loss": 0.0154,
      "step": 4900
    },
    {
      "epoch": 3.236651285431773,
      "grad_norm": 0.2565387487411499,
      "learning_rate": 1.0036913034775674e-05,
      "loss": 0.0289,
      "step": 4910
    },
    {
      "epoch": 3.2432432432432434,
      "grad_norm": 0.33470308780670166,
      "learning_rate": 9.784496459457549e-06,
      "loss": 0.0178,
      "step": 4920
    },
    {
      "epoch": 3.2498352010547134,
      "grad_norm": 0.2145528793334961,
      "learning_rate": 9.535131083180027e-06,
      "loss": 0.0225,
      "step": 4930
    },
    {
      "epoch": 3.2564271588661833,
      "grad_norm": 0.2850882411003113,
      "learning_rate": 9.288825339786133e-06,
      "loss": 0.0212,
      "step": 4940
    },
    {
      "epoch": 3.263019116677653,
      "grad_norm": 0.30832791328430176,
      "learning_rate": 9.045587559638358e-06,
      "loss": 0.0179,
      "step": 4950
    },
    {
      "epoch": 3.269611074489123,
      "grad_norm": 0.301908403635025,
      "learning_rate": 8.805425969336945e-06,
      "loss": 0.02,
      "step": 4960
    },
    {
      "epoch": 3.276203032300593,
      "grad_norm": 0.24023061990737915,
      "learning_rate": 8.568348691441596e-06,
      "loss": 0.0188,
      "step": 4970
    },
    {
      "epoch": 3.2827949901120634,
      "grad_norm": 0.22343115508556366,
      "learning_rate": 8.33436374419676e-06,
      "loss": 0.0215,
      "step": 4980
    },
    {
      "epoch": 3.2893869479235334,
      "grad_norm": 0.29182159900665283,
      "learning_rate": 8.10347904126053e-06,
      "loss": 0.0229,
      "step": 4990
    },
    {
      "epoch": 3.2959789057350033,
      "grad_norm": 0.2520536780357361,
      "learning_rate": 7.87570239143689e-06,
      "loss": 0.0219,
      "step": 5000
    },
    {
      "epoch": 3.3025708635464732,
      "grad_norm": 0.3384230136871338,
      "learning_rate": 7.651041498411637e-06,
      "loss": 0.0199,
      "step": 5010
    },
    {
      "epoch": 3.309162821357943,
      "grad_norm": 0.30371981859207153,
      "learning_rate": 7.429503960491901e-06,
      "loss": 0.0231,
      "step": 5020
    },
    {
      "epoch": 3.3157547791694135,
      "grad_norm": 0.21716776490211487,
      "learning_rate": 7.211097270349066e-06,
      "loss": 0.0158,
      "step": 5030
    },
    {
      "epoch": 3.3223467369808835,
      "grad_norm": 0.2968665659427643,
      "learning_rate": 6.995828814765426e-06,
      "loss": 0.0153,
      "step": 5040
    },
    {
      "epoch": 3.3289386947923534,
      "grad_norm": 0.2764769196510315,
      "learning_rate": 6.7837058743843275e-06,
      "loss": 0.0194,
      "step": 5050
    },
    {
      "epoch": 3.3355306526038233,
      "grad_norm": 0.3363369405269623,
      "learning_rate": 6.57473562346389e-06,
      "loss": 0.0156,
      "step": 5060
    },
    {
      "epoch": 3.3421226104152932,
      "grad_norm": 0.386843740940094,
      "learning_rate": 6.3689251296344845e-06,
      "loss": 0.0208,
      "step": 5070
    },
    {
      "epoch": 3.348714568226763,
      "grad_norm": 0.142068549990654,
      "learning_rate": 6.166281353659575e-06,
      "loss": 0.0205,
      "step": 5080
    },
    {
      "epoch": 3.3553065260382335,
      "grad_norm": 0.4907970130443573,
      "learning_rate": 5.966811149200324e-06,
      "loss": 0.0201,
      "step": 5090
    },
    {
      "epoch": 3.3618984838497035,
      "grad_norm": 0.3856694996356964,
      "learning_rate": 5.77052126258385e-06,
      "loss": 0.0259,
      "step": 5100
    },
    {
      "epoch": 3.3684904416611734,
      "grad_norm": 0.28562742471694946,
      "learning_rate": 5.577418332575002e-06,
      "loss": 0.02,
      "step": 5110
    },
    {
      "epoch": 3.3750823994726433,
      "grad_norm": 0.24548883736133575,
      "learning_rate": 5.387508890151827e-06,
      "loss": 0.0194,
      "step": 5120
    },
    {
      "epoch": 3.3816743572841133,
      "grad_norm": 0.19759061932563782,
      "learning_rate": 5.200799358284758e-06,
      "loss": 0.0148,
      "step": 5130
    },
    {
      "epoch": 3.388266315095583,
      "grad_norm": 0.41944679617881775,
      "learning_rate": 5.017296051719256e-06,
      "loss": 0.0199,
      "step": 5140
    },
    {
      "epoch": 3.3948582729070536,
      "grad_norm": 0.42225712537765503,
      "learning_rate": 4.837005176762355e-06,
      "loss": 0.0225,
      "step": 5150
    },
    {
      "epoch": 3.4014502307185235,
      "grad_norm": 0.2787589728832245,
      "learning_rate": 4.659932831072677e-06,
      "loss": 0.0189,
      "step": 5160
    },
    {
      "epoch": 3.4080421885299934,
      "grad_norm": 0.3040010333061218,
      "learning_rate": 4.4860850034542765e-06,
      "loss": 0.0197,
      "step": 5170
    },
    {
      "epoch": 3.4146341463414633,
      "grad_norm": 0.23647019267082214,
      "learning_rate": 4.315467573654031e-06,
      "loss": 0.0201,
      "step": 5180
    },
    {
      "epoch": 3.4212261041529333,
      "grad_norm": 0.3153569996356964,
      "learning_rate": 4.148086312162758e-06,
      "loss": 0.0201,
      "step": 5190
    },
    {
      "epoch": 3.4278180619644036,
      "grad_norm": 0.37144842743873596,
      "learning_rate": 3.983946880020173e-06,
      "loss": 0.0192,
      "step": 5200
    },
    {
      "epoch": 3.4344100197758736,
      "grad_norm": 0.30148690938949585,
      "learning_rate": 3.823054828623285e-06,
      "loss": 0.0196,
      "step": 5210
    },
    {
      "epoch": 3.4410019775873435,
      "grad_norm": 0.2592622935771942,
      "learning_rate": 3.6654155995386905e-06,
      "loss": 0.0145,
      "step": 5220
    },
    {
      "epoch": 3.4475939353988134,
      "grad_norm": 0.45873746275901794,
      "learning_rate": 3.5110345243186015e-06,
      "loss": 0.0233,
      "step": 5230
    },
    {
      "epoch": 3.4541858932102834,
      "grad_norm": 0.2648702561855316,
      "learning_rate": 3.3599168243204193e-06,
      "loss": 0.0205,
      "step": 5240
    },
    {
      "epoch": 3.4607778510217533,
      "grad_norm": 0.3078472316265106,
      "learning_rate": 3.212067610530212e-06,
      "loss": 0.0151,
      "step": 5250
    },
    {
      "epoch": 3.4673698088332237,
      "grad_norm": 0.237100750207901,
      "learning_rate": 3.067491883389839e-06,
      "loss": 0.0165,
      "step": 5260
    },
    {
      "epoch": 3.4739617666446936,
      "grad_norm": 0.21496574580669403,
      "learning_rate": 2.9261945326277794e-06,
      "loss": 0.0165,
      "step": 5270
    },
    {
      "epoch": 3.4805537244561635,
      "grad_norm": 0.24198366701602936,
      "learning_rate": 2.7881803370938597e-06,
      "loss": 0.0171,
      "step": 5280
    },
    {
      "epoch": 3.4871456822676334,
      "grad_norm": 0.20283302664756775,
      "learning_rate": 2.653453964597563e-06,
      "loss": 0.0148,
      "step": 5290
    },
    {
      "epoch": 3.4937376400791034,
      "grad_norm": 0.24728865921497345,
      "learning_rate": 2.522019971750089e-06,
      "loss": 0.0208,
      "step": 5300
    },
    {
      "epoch": 3.5003295978905733,
      "grad_norm": 0.18928475677967072,
      "learning_rate": 2.393882803810388e-06,
      "loss": 0.0305,
      "step": 5310
    },
    {
      "epoch": 3.5069215557020437,
      "grad_norm": 0.31071001291275024,
      "learning_rate": 2.2690467945347682e-06,
      "loss": 0.019,
      "step": 5320
    },
    {
      "epoch": 3.5135135135135136,
      "grad_norm": 0.49363845586776733,
      "learning_rate": 2.1475161660301944e-06,
      "loss": 0.0217,
      "step": 5330
    },
    {
      "epoch": 3.5201054713249835,
      "grad_norm": 0.464982807636261,
      "learning_rate": 2.0292950286116996e-06,
      "loss": 0.0167,
      "step": 5340
    },
    {
      "epoch": 3.5266974291364535,
      "grad_norm": 0.5095881819725037,
      "learning_rate": 1.9143873806632406e-06,
      "loss": 0.0221,
      "step": 5350
    },
    {
      "epoch": 3.5332893869479234,
      "grad_norm": 0.23975181579589844,
      "learning_rate": 1.802797108502452e-06,
      "loss": 0.0226,
      "step": 5360
    },
    {
      "epoch": 3.5398813447593938,
      "grad_norm": 0.30347514152526855,
      "learning_rate": 1.6945279862493059e-06,
      "loss": 0.0187,
      "step": 5370
    },
    {
      "epoch": 3.5464733025708637,
      "grad_norm": 0.29192495346069336,
      "learning_rate": 1.5895836756983695e-06,
      "loss": 0.0202,
      "step": 5380
    },
    {
      "epoch": 3.5530652603823336,
      "grad_norm": 0.2690478563308716,
      "learning_rate": 1.4879677261950164e-06,
      "loss": 0.0225,
      "step": 5390
    },
    {
      "epoch": 3.5596572181938035,
      "grad_norm": 0.2959778904914856,
      "learning_rate": 1.3896835745153547e-06,
      "loss": 0.0204,
      "step": 5400
    },
    {
      "epoch": 3.5662491760052735,
      "grad_norm": 0.31142422556877136,
      "learning_rate": 1.2947345447500092e-06,
      "loss": 0.0242,
      "step": 5410
    },
    {
      "epoch": 3.572841133816744,
      "grad_norm": 0.1820339411497116,
      "learning_rate": 1.2031238481916896e-06,
      "loss": 0.0213,
      "step": 5420
    },
    {
      "epoch": 3.5794330916282133,
      "grad_norm": 0.24110187590122223,
      "learning_rate": 1.1148545832265767e-06,
      "loss": 0.0172,
      "step": 5430
    },
    {
      "epoch": 3.5860250494396837,
      "grad_norm": 0.14942501485347748,
      "learning_rate": 1.029929735229529e-06,
      "loss": 0.0204,
      "step": 5440
    },
    {
      "epoch": 3.5926170072511536,
      "grad_norm": 0.4690724313259125,
      "learning_rate": 9.483521764631298e-07,
      "loss": 0.0219,
      "step": 5450
    },
    {
      "epoch": 3.5992089650626236,
      "grad_norm": 0.3038761019706726,
      "learning_rate": 8.701246659805206e-07,
      "loss": 0.0134,
      "step": 5460
    },
    {
      "epoch": 3.6058009228740935,
      "grad_norm": 0.2972651422023773,
      "learning_rate": 7.952498495321093e-07,
      "loss": 0.0191,
      "step": 5470
    },
    {
      "epoch": 3.6123928806855634,
      "grad_norm": 0.4311395287513733,
      "learning_rate": 7.237302594760741e-07,
      "loss": 0.0208,
      "step": 5480
    },
    {
      "epoch": 3.618984838497034,
      "grad_norm": 0.45453134179115295,
      "learning_rate": 6.555683146927117e-07,
      "loss": 0.0158,
      "step": 5490
    },
    {
      "epoch": 3.6255767963085037,
      "grad_norm": 0.31247788667678833,
      "learning_rate": 5.907663205026559e-07,
      "loss": 0.0191,
      "step": 5500
    }
  ],
  "logging_steps": 10,
  "max_steps": 5687,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.928222264873016e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}