{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0002680725583057,
  "eval_steps": 500,
  "global_step": 11194,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008935751943526047,
      "grad_norm": 4.83393669128418,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 1.4311,
      "step": 10
    },
    {
      "epoch": 0.0017871503887052094,
      "grad_norm": 3.230992555618286,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 1.4013,
      "step": 20
    },
    {
      "epoch": 0.0026807255830578144,
      "grad_norm": 1.9909214973449707,
      "learning_rate": 3.342134831460675e-06,
      "loss": 1.1381,
      "step": 30
    },
    {
      "epoch": 0.003574300777410419,
      "grad_norm": 0.8147533535957336,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 0.9713,
      "step": 40
    },
    {
      "epoch": 0.004467875971763024,
      "grad_norm": 0.8770450949668884,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 0.8768,
      "step": 50
    },
    {
      "epoch": 0.005361451166115629,
      "grad_norm": 0.4910338222980499,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 0.8431,
      "step": 60
    },
    {
      "epoch": 0.006255026360468234,
      "grad_norm": 0.5245729684829712,
      "learning_rate": 7.814044943820226e-06,
      "loss": 0.7821,
      "step": 70
    },
    {
      "epoch": 0.007148601554820838,
      "grad_norm": 0.4985717535018921,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.8137,
      "step": 80
    },
    {
      "epoch": 0.008042176749173444,
      "grad_norm": 0.4982825517654419,
      "learning_rate": 1.005e-05,
      "loss": 0.8014,
      "step": 90
    },
    {
      "epoch": 0.008935751943526048,
      "grad_norm": 0.4962609112262726,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.8102,
      "step": 100
    },
    {
      "epoch": 0.009829327137878652,
      "grad_norm": 0.4608575701713562,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.7836,
      "step": 110
    },
    {
      "epoch": 0.010722902332231258,
      "grad_norm": 0.4405043423175812,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.7987,
      "step": 120
    },
    {
      "epoch": 0.011616477526583862,
      "grad_norm": 0.4832342565059662,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.757,
      "step": 130
    },
    {
      "epoch": 0.012510052720936467,
      "grad_norm": 0.5148149132728577,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.7922,
      "step": 140
    },
    {
      "epoch": 0.013403627915289072,
      "grad_norm": 0.496463418006897,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.8236,
      "step": 150
    },
    {
      "epoch": 0.014297203109641676,
      "grad_norm": 0.5295011401176453,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.798,
      "step": 160
    },
    {
      "epoch": 0.015190778303994281,
      "grad_norm": 0.47589513659477234,
      "learning_rate": 1.89938202247191e-05,
      "loss": 0.8071,
      "step": 170
    },
    {
      "epoch": 0.016084353498346887,
      "grad_norm": 0.42235517501831055,
      "learning_rate": 2e-05,
      "loss": 0.7128,
      "step": 180
    },
    {
      "epoch": 0.01697792869269949,
      "grad_norm": 0.4666895270347595,
      "learning_rate": 2e-05,
      "loss": 0.7695,
      "step": 190
    },
    {
      "epoch": 0.017871503887052095,
      "grad_norm": 0.41975900530815125,
      "learning_rate": 2e-05,
      "loss": 0.7967,
      "step": 200
    },
    {
      "epoch": 0.0187650790814047,
      "grad_norm": 0.4837177097797394,
      "learning_rate": 2e-05,
      "loss": 0.7712,
      "step": 210
    },
    {
      "epoch": 0.019658654275757304,
      "grad_norm": 0.422789603471756,
      "learning_rate": 2e-05,
      "loss": 0.7704,
      "step": 220
    },
    {
      "epoch": 0.02055222947010991,
      "grad_norm": 0.42405927181243896,
      "learning_rate": 2e-05,
      "loss": 0.7445,
      "step": 230
    },
    {
      "epoch": 0.021445804664462515,
      "grad_norm": 0.44666311144828796,
      "learning_rate": 2e-05,
      "loss": 0.7747,
      "step": 240
    },
    {
      "epoch": 0.02233937985881512,
      "grad_norm": 0.413830041885376,
      "learning_rate": 2e-05,
      "loss": 0.7766,
      "step": 250
    },
    {
      "epoch": 0.023232955053167723,
      "grad_norm": 0.4343000054359436,
      "learning_rate": 2e-05,
      "loss": 0.8023,
      "step": 260
    },
    {
      "epoch": 0.024126530247520327,
      "grad_norm": 0.4592699110507965,
      "learning_rate": 2e-05,
      "loss": 0.7824,
      "step": 270
    },
    {
      "epoch": 0.025020105441872935,
      "grad_norm": 0.4644310474395752,
      "learning_rate": 2e-05,
      "loss": 0.7371,
      "step": 280
    },
    {
      "epoch": 0.02591368063622554,
      "grad_norm": 0.4319989085197449,
      "learning_rate": 2e-05,
      "loss": 0.7641,
      "step": 290
    },
    {
      "epoch": 0.026807255830578143,
      "grad_norm": 0.4405927360057831,
      "learning_rate": 2e-05,
      "loss": 0.7863,
      "step": 300
    },
    {
      "epoch": 0.027700831024930747,
      "grad_norm": 0.4252799451351166,
      "learning_rate": 2e-05,
      "loss": 0.7928,
      "step": 310
    },
    {
      "epoch": 0.02859440621928335,
      "grad_norm": 0.4097546637058258,
      "learning_rate": 2e-05,
      "loss": 0.7965,
      "step": 320
    },
    {
      "epoch": 0.02948798141363596,
      "grad_norm": 0.4354838728904724,
      "learning_rate": 2e-05,
      "loss": 0.7932,
      "step": 330
    },
    {
      "epoch": 0.030381556607988563,
      "grad_norm": 0.4520812928676605,
      "learning_rate": 2e-05,
      "loss": 0.793,
      "step": 340
    },
    {
      "epoch": 0.03127513180234117,
      "grad_norm": 0.5291610360145569,
      "learning_rate": 2e-05,
      "loss": 0.7466,
      "step": 350
    },
    {
      "epoch": 0.032168706996693774,
      "grad_norm": 0.442043662071228,
      "learning_rate": 2e-05,
      "loss": 0.7455,
      "step": 360
    },
    {
      "epoch": 0.03306228219104638,
      "grad_norm": 0.39435988664627075,
      "learning_rate": 2e-05,
      "loss": 0.7664,
      "step": 370
    },
    {
      "epoch": 0.03395585738539898,
      "grad_norm": 0.6453039050102234,
      "learning_rate": 2e-05,
      "loss": 0.7415,
      "step": 380
    },
    {
      "epoch": 0.03484943257975159,
      "grad_norm": 0.6479556560516357,
      "learning_rate": 2e-05,
      "loss": 0.8238,
      "step": 390
    },
    {
      "epoch": 0.03574300777410419,
      "grad_norm": 0.4728488624095917,
      "learning_rate": 2e-05,
      "loss": 0.7672,
      "step": 400
    },
    {
      "epoch": 0.036636582968456795,
      "grad_norm": 0.4299718141555786,
      "learning_rate": 2e-05,
      "loss": 0.7397,
      "step": 410
    },
    {
      "epoch": 0.0375301581628094,
      "grad_norm": 0.35715794563293457,
      "learning_rate": 2e-05,
      "loss": 0.7332,
      "step": 420
    },
    {
      "epoch": 0.038423733357162,
      "grad_norm": 0.430271178483963,
      "learning_rate": 2e-05,
      "loss": 0.7557,
      "step": 430
    },
    {
      "epoch": 0.03931730855151461,
      "grad_norm": 0.4169669449329376,
      "learning_rate": 2e-05,
      "loss": 0.7654,
      "step": 440
    },
    {
      "epoch": 0.04021088374586722,
      "grad_norm": 0.4232149124145508,
      "learning_rate": 2e-05,
      "loss": 0.7845,
      "step": 450
    },
    {
      "epoch": 0.04110445894021982,
      "grad_norm": 0.42479565739631653,
      "learning_rate": 2e-05,
      "loss": 0.7642,
      "step": 460
    },
    {
      "epoch": 0.041998034134572426,
      "grad_norm": 0.38824278116226196,
      "learning_rate": 2e-05,
      "loss": 0.7514,
      "step": 470
    },
    {
      "epoch": 0.04289160932892503,
      "grad_norm": 0.4207800626754761,
      "learning_rate": 2e-05,
      "loss": 0.7528,
      "step": 480
    },
    {
      "epoch": 0.043785184523277634,
      "grad_norm": 0.40183427929878235,
      "learning_rate": 2e-05,
      "loss": 0.7806,
      "step": 490
    },
    {
      "epoch": 0.04467875971763024,
      "grad_norm": 0.431792676448822,
      "learning_rate": 2e-05,
      "loss": 0.7755,
      "step": 500
    },
    {
      "epoch": 0.04557233491198284,
      "grad_norm": 0.4143838584423065,
      "learning_rate": 2e-05,
      "loss": 0.7539,
      "step": 510
    },
    {
      "epoch": 0.04646591010633545,
      "grad_norm": 0.44914576411247253,
      "learning_rate": 2e-05,
      "loss": 0.7638,
      "step": 520
    },
    {
      "epoch": 0.04735948530068805,
      "grad_norm": 0.3990241587162018,
      "learning_rate": 2e-05,
      "loss": 0.7566,
      "step": 530
    },
    {
      "epoch": 0.048253060495040655,
      "grad_norm": 0.4371657371520996,
      "learning_rate": 2e-05,
      "loss": 0.7551,
      "step": 540
    },
    {
      "epoch": 0.049146635689393266,
      "grad_norm": 0.45287543535232544,
      "learning_rate": 2e-05,
      "loss": 0.7469,
      "step": 550
    },
    {
      "epoch": 0.05004021088374587,
      "grad_norm": 0.3757151961326599,
      "learning_rate": 2e-05,
      "loss": 0.7494,
      "step": 560
    },
    {
      "epoch": 0.050933786078098474,
      "grad_norm": 0.5098564624786377,
      "learning_rate": 2e-05,
      "loss": 0.7648,
      "step": 570
    },
    {
      "epoch": 0.05182736127245108,
      "grad_norm": 0.43075209856033325,
      "learning_rate": 2e-05,
      "loss": 0.7296,
      "step": 580
    },
    {
      "epoch": 0.05272093646680368,
      "grad_norm": 0.46172618865966797,
      "learning_rate": 2e-05,
      "loss": 0.7439,
      "step": 590
    },
    {
      "epoch": 0.053614511661156286,
      "grad_norm": 0.4297853112220764,
      "learning_rate": 2e-05,
      "loss": 0.7748,
      "step": 600
    },
    {
      "epoch": 0.05450808685550889,
      "grad_norm": 0.39287275075912476,
      "learning_rate": 2e-05,
      "loss": 0.7353,
      "step": 610
    },
    {
      "epoch": 0.055401662049861494,
      "grad_norm": 0.3974439203739166,
      "learning_rate": 2e-05,
      "loss": 0.7451,
      "step": 620
    },
    {
      "epoch": 0.0562952372442141,
      "grad_norm": 0.39414021372795105,
      "learning_rate": 2e-05,
      "loss": 0.7739,
      "step": 630
    },
    {
      "epoch": 0.0571888124385667,
      "grad_norm": 0.42754384875297546,
      "learning_rate": 2e-05,
      "loss": 0.779,
      "step": 640
    },
    {
      "epoch": 0.05808238763291931,
      "grad_norm": 0.3786719739437103,
      "learning_rate": 2e-05,
      "loss": 0.7636,
      "step": 650
    },
    {
      "epoch": 0.05897596282727192,
      "grad_norm": 0.36745455861091614,
      "learning_rate": 2e-05,
      "loss": 0.7749,
      "step": 660
    },
    {
      "epoch": 0.05986953802162452,
      "grad_norm": 0.44898897409439087,
      "learning_rate": 2e-05,
      "loss": 0.7306,
      "step": 670
    },
    {
      "epoch": 0.060763113215977126,
      "grad_norm": 0.4336882531642914,
      "learning_rate": 2e-05,
      "loss": 0.7385,
      "step": 680
    },
    {
      "epoch": 0.06165668841032973,
      "grad_norm": 0.3571746349334717,
      "learning_rate": 2e-05,
      "loss": 0.7209,
      "step": 690
    },
    {
      "epoch": 0.06255026360468234,
      "grad_norm": 0.4294300079345703,
      "learning_rate": 2e-05,
      "loss": 0.7694,
      "step": 700
    },
    {
      "epoch": 0.06344383879903494,
      "grad_norm": 0.3911939859390259,
      "learning_rate": 2e-05,
      "loss": 0.7571,
      "step": 710
    },
    {
      "epoch": 0.06433741399338755,
      "grad_norm": 0.44994938373565674,
      "learning_rate": 2e-05,
      "loss": 0.7509,
      "step": 720
    },
    {
      "epoch": 0.06523098918774015,
      "grad_norm": 0.43563222885131836,
      "learning_rate": 2e-05,
      "loss": 0.7727,
      "step": 730
    },
    {
      "epoch": 0.06612456438209276,
      "grad_norm": 0.4137369394302368,
      "learning_rate": 2e-05,
      "loss": 0.7598,
      "step": 740
    },
    {
      "epoch": 0.06701813957644535,
      "grad_norm": 0.3853653073310852,
      "learning_rate": 2e-05,
      "loss": 0.7585,
      "step": 750
    },
    {
      "epoch": 0.06791171477079797,
      "grad_norm": 0.3681003153324127,
      "learning_rate": 2e-05,
      "loss": 0.7394,
      "step": 760
    },
    {
      "epoch": 0.06880528996515056,
      "grad_norm": 0.3928927183151245,
      "learning_rate": 2e-05,
      "loss": 0.7484,
      "step": 770
    },
    {
      "epoch": 0.06969886515950317,
      "grad_norm": 0.387007474899292,
      "learning_rate": 2e-05,
      "loss": 0.7262,
      "step": 780
    },
    {
      "epoch": 0.07059244035385577,
      "grad_norm": 0.46645933389663696,
      "learning_rate": 2e-05,
      "loss": 0.7281,
      "step": 790
    },
    {
      "epoch": 0.07148601554820838,
      "grad_norm": 0.4278911352157593,
      "learning_rate": 2e-05,
      "loss": 0.7226,
      "step": 800
    },
    {
      "epoch": 0.07237959074256099,
      "grad_norm": 0.39950302243232727,
      "learning_rate": 2e-05,
      "loss": 0.7012,
      "step": 810
    },
    {
      "epoch": 0.07327316593691359,
      "grad_norm": 0.3807908892631531,
      "learning_rate": 2e-05,
      "loss": 0.7324,
      "step": 820
    },
    {
      "epoch": 0.0741667411312662,
      "grad_norm": 0.4291982352733612,
      "learning_rate": 2e-05,
      "loss": 0.7244,
      "step": 830
    },
    {
      "epoch": 0.0750603163256188,
      "grad_norm": 0.40053242444992065,
      "learning_rate": 2e-05,
      "loss": 0.7638,
      "step": 840
    },
    {
      "epoch": 0.07595389151997141,
      "grad_norm": 0.4022897481918335,
      "learning_rate": 2e-05,
      "loss": 0.753,
      "step": 850
    },
    {
      "epoch": 0.076847466714324,
      "grad_norm": 0.5166751146316528,
      "learning_rate": 2e-05,
      "loss": 0.7418,
      "step": 860
    },
    {
      "epoch": 0.07774104190867662,
      "grad_norm": 0.4001097083091736,
      "learning_rate": 2e-05,
      "loss": 0.7557,
      "step": 870
    },
    {
      "epoch": 0.07863461710302921,
      "grad_norm": 0.37985655665397644,
      "learning_rate": 2e-05,
      "loss": 0.7429,
      "step": 880
    },
    {
      "epoch": 0.07952819229738183,
      "grad_norm": 0.3874037563800812,
      "learning_rate": 2e-05,
      "loss": 0.7409,
      "step": 890
    },
    {
      "epoch": 0.08042176749173444,
      "grad_norm": 0.40059277415275574,
      "learning_rate": 2e-05,
      "loss": 0.7468,
      "step": 900
    },
    {
      "epoch": 0.08131534268608703,
      "grad_norm": 0.3915744423866272,
      "learning_rate": 2e-05,
      "loss": 0.7425,
      "step": 910
    },
    {
      "epoch": 0.08220891788043964,
      "grad_norm": 0.4296461343765259,
      "learning_rate": 2e-05,
      "loss": 0.7538,
      "step": 920
    },
    {
      "epoch": 0.08310249307479224,
      "grad_norm": 0.39975130558013916,
      "learning_rate": 2e-05,
      "loss": 0.7751,
      "step": 930
    },
    {
      "epoch": 0.08399606826914485,
      "grad_norm": 0.4234040379524231,
      "learning_rate": 2e-05,
      "loss": 0.7217,
      "step": 940
    },
    {
      "epoch": 0.08488964346349745,
      "grad_norm": 0.7239550352096558,
      "learning_rate": 2e-05,
      "loss": 0.7523,
      "step": 950
    },
    {
      "epoch": 0.08578321865785006,
      "grad_norm": 0.40926212072372437,
      "learning_rate": 2e-05,
      "loss": 0.7073,
      "step": 960
    },
    {
      "epoch": 0.08667679385220266,
      "grad_norm": 0.42253273725509644,
      "learning_rate": 2e-05,
      "loss": 0.7645,
      "step": 970
    },
    {
      "epoch": 0.08757036904655527,
      "grad_norm": 0.42166000604629517,
      "learning_rate": 2e-05,
      "loss": 0.7206,
      "step": 980
    },
    {
      "epoch": 0.08846394424090787,
      "grad_norm": 0.4257868826389313,
      "learning_rate": 2e-05,
      "loss": 0.7336,
      "step": 990
    },
    {
      "epoch": 0.08935751943526048,
      "grad_norm": 0.3908492922782898,
      "learning_rate": 2e-05,
      "loss": 0.7297,
      "step": 1000
    },
    {
      "epoch": 0.09025109462961309,
      "grad_norm": 0.38620543479919434,
      "learning_rate": 2e-05,
      "loss": 0.7451,
      "step": 1010
    },
    {
      "epoch": 0.09114466982396568,
      "grad_norm": 0.40934133529663086,
      "learning_rate": 2e-05,
      "loss": 0.7403,
      "step": 1020
    },
    {
      "epoch": 0.0920382450183183,
      "grad_norm": 0.435519814491272,
      "learning_rate": 2e-05,
      "loss": 0.7144,
      "step": 1030
    },
    {
      "epoch": 0.0929318202126709,
      "grad_norm": 0.46169203519821167,
      "learning_rate": 2e-05,
      "loss": 0.7317,
      "step": 1040
    },
    {
      "epoch": 0.0938253954070235,
      "grad_norm": 0.37340661883354187,
      "learning_rate": 2e-05,
      "loss": 0.7714,
      "step": 1050
    },
    {
      "epoch": 0.0947189706013761,
      "grad_norm": 0.41289299726486206,
      "learning_rate": 2e-05,
      "loss": 0.7351,
      "step": 1060
    },
    {
      "epoch": 0.09561254579572871,
      "grad_norm": 0.4187810719013214,
      "learning_rate": 2e-05,
      "loss": 0.7327,
      "step": 1070
    },
    {
      "epoch": 0.09650612099008131,
      "grad_norm": 0.41718336939811707,
      "learning_rate": 2e-05,
      "loss": 0.7864,
      "step": 1080
    },
    {
      "epoch": 0.09739969618443392,
      "grad_norm": 0.4145309329032898,
      "learning_rate": 2e-05,
      "loss": 0.7505,
      "step": 1090
    },
    {
      "epoch": 0.09829327137878653,
      "grad_norm": 0.4168350100517273,
      "learning_rate": 2e-05,
      "loss": 0.7359,
      "step": 1100
    },
    {
      "epoch": 0.09918684657313913,
      "grad_norm": 0.39129555225372314,
      "learning_rate": 2e-05,
      "loss": 0.7321,
      "step": 1110
    },
    {
      "epoch": 0.10008042176749174,
      "grad_norm": 0.39075788855552673,
      "learning_rate": 2e-05,
      "loss": 0.7625,
      "step": 1120
    },
    {
      "epoch": 0.10097399696184434,
      "grad_norm": 0.4408645033836365,
      "learning_rate": 2e-05,
      "loss": 0.7388,
      "step": 1130
    },
    {
      "epoch": 0.10186757215619695,
      "grad_norm": 0.3980259597301483,
      "learning_rate": 2e-05,
      "loss": 0.7537,
      "step": 1140
    },
    {
      "epoch": 0.10276114735054954,
      "grad_norm": 0.3763454556465149,
      "learning_rate": 2e-05,
      "loss": 0.7488,
      "step": 1150
    },
    {
      "epoch": 0.10365472254490216,
      "grad_norm": 0.37730491161346436,
      "learning_rate": 2e-05,
      "loss": 0.7465,
      "step": 1160
    },
    {
      "epoch": 0.10454829773925475,
      "grad_norm": 0.37538275122642517,
      "learning_rate": 2e-05,
      "loss": 0.7365,
      "step": 1170
    },
    {
      "epoch": 0.10544187293360736,
      "grad_norm": 0.397089421749115,
      "learning_rate": 2e-05,
      "loss": 0.769,
      "step": 1180
    },
    {
      "epoch": 0.10633544812795996,
      "grad_norm": 0.3705765902996063,
      "learning_rate": 2e-05,
      "loss": 0.7385,
      "step": 1190
    },
    {
      "epoch": 0.10722902332231257,
      "grad_norm": 0.3545670807361603,
      "learning_rate": 2e-05,
      "loss": 0.7408,
      "step": 1200
    },
    {
      "epoch": 0.10812259851666518,
      "grad_norm": 0.4176211357116699,
      "learning_rate": 2e-05,
      "loss": 0.7409,
      "step": 1210
    },
    {
      "epoch": 0.10901617371101778,
      "grad_norm": 0.36956384778022766,
      "learning_rate": 2e-05,
      "loss": 0.7246,
      "step": 1220
    },
    {
      "epoch": 0.10990974890537039,
      "grad_norm": 0.3938690423965454,
      "learning_rate": 2e-05,
      "loss": 0.7292,
      "step": 1230
    },
    {
      "epoch": 0.11080332409972299,
      "grad_norm": 0.39145007729530334,
      "learning_rate": 2e-05,
      "loss": 0.7285,
      "step": 1240
    },
    {
      "epoch": 0.1116968992940756,
      "grad_norm": 0.38112127780914307,
      "learning_rate": 2e-05,
      "loss": 0.7551,
      "step": 1250
    },
    {
      "epoch": 0.1125904744884282,
      "grad_norm": 0.3316250741481781,
      "learning_rate": 2e-05,
      "loss": 0.7323,
      "step": 1260
    },
    {
      "epoch": 0.11348404968278081,
      "grad_norm": 0.4049963057041168,
      "learning_rate": 2e-05,
      "loss": 0.7329,
      "step": 1270
    },
    {
      "epoch": 0.1143776248771334,
      "grad_norm": 0.48271071910858154,
      "learning_rate": 2e-05,
      "loss": 0.7696,
      "step": 1280
    },
    {
      "epoch": 0.11527120007148602,
      "grad_norm": 0.37332484126091003,
      "learning_rate": 2e-05,
      "loss": 0.7045,
      "step": 1290
    },
    {
      "epoch": 0.11616477526583863,
      "grad_norm": 0.3976444602012634,
      "learning_rate": 2e-05,
      "loss": 0.7259,
      "step": 1300
    },
    {
      "epoch": 0.11705835046019122,
      "grad_norm": 0.39804327487945557,
      "learning_rate": 2e-05,
      "loss": 0.7406,
      "step": 1310
    },
    {
      "epoch": 0.11795192565454383,
      "grad_norm": 0.4065830707550049,
      "learning_rate": 2e-05,
      "loss": 0.7332,
      "step": 1320
    },
    {
      "epoch": 0.11884550084889643,
      "grad_norm": 0.38445714116096497,
      "learning_rate": 2e-05,
      "loss": 0.7277,
      "step": 1330
    },
    {
      "epoch": 0.11973907604324904,
      "grad_norm": 0.3987727761268616,
      "learning_rate": 2e-05,
      "loss": 0.6999,
      "step": 1340
    },
    {
      "epoch": 0.12063265123760164,
      "grad_norm": 0.37083423137664795,
      "learning_rate": 2e-05,
      "loss": 0.7425,
      "step": 1350
    },
    {
      "epoch": 0.12152622643195425,
      "grad_norm": 0.4784857928752899,
      "learning_rate": 2e-05,
      "loss": 0.7522,
      "step": 1360
    },
    {
      "epoch": 0.12241980162630685,
      "grad_norm": 0.39088118076324463,
      "learning_rate": 2e-05,
      "loss": 0.7366,
      "step": 1370
    },
    {
      "epoch": 0.12331337682065946,
      "grad_norm": 0.39330583810806274,
      "learning_rate": 2e-05,
      "loss": 0.7293,
      "step": 1380
    },
    {
      "epoch": 0.12420695201501206,
      "grad_norm": 0.3860457241535187,
      "learning_rate": 2e-05,
      "loss": 0.7143,
      "step": 1390
    },
    {
      "epoch": 0.12510052720936468,
      "grad_norm": 0.38415059447288513,
      "learning_rate": 2e-05,
      "loss": 0.6834,
      "step": 1400
    },
    {
      "epoch": 0.12599410240371728,
      "grad_norm": 0.39965566992759705,
      "learning_rate": 2e-05,
      "loss": 0.7087,
      "step": 1410
    },
    {
      "epoch": 0.12688767759806988,
      "grad_norm": 0.38492318987846375,
      "learning_rate": 2e-05,
      "loss": 0.7547,
      "step": 1420
    },
    {
      "epoch": 0.12778125279242247,
      "grad_norm": 0.41795384883880615,
      "learning_rate": 2e-05,
      "loss": 0.7267,
      "step": 1430
    },
    {
      "epoch": 0.1286748279867751,
      "grad_norm": 0.3837740123271942,
      "learning_rate": 2e-05,
      "loss": 0.74,
      "step": 1440
    },
    {
      "epoch": 0.1295684031811277,
      "grad_norm": 0.35027605295181274,
      "learning_rate": 2e-05,
      "loss": 0.6958,
      "step": 1450
    },
    {
      "epoch": 0.1304619783754803,
      "grad_norm": 0.36292174458503723,
      "learning_rate": 2e-05,
      "loss": 0.7441,
      "step": 1460
    },
    {
      "epoch": 0.1313555535698329,
      "grad_norm": 0.3984721899032593,
      "learning_rate": 2e-05,
      "loss": 0.7197,
      "step": 1470
    },
    {
      "epoch": 0.13224912876418551,
      "grad_norm": 0.4132763147354126,
      "learning_rate": 2e-05,
      "loss": 0.7457,
      "step": 1480
    },
    {
      "epoch": 0.1331427039585381,
      "grad_norm": 0.3790658712387085,
      "learning_rate": 2e-05,
      "loss": 0.7274,
      "step": 1490
    },
    {
      "epoch": 0.1340362791528907,
      "grad_norm": 0.3779606819152832,
      "learning_rate": 2e-05,
      "loss": 0.7259,
      "step": 1500
    },
    {
      "epoch": 0.13492985434724333,
      "grad_norm": 0.9285029172897339,
      "learning_rate": 2e-05,
      "loss": 0.7231,
      "step": 1510
    },
    {
      "epoch": 0.13582342954159593,
      "grad_norm": 0.3995453119277954,
      "learning_rate": 2e-05,
      "loss": 0.7022,
      "step": 1520
    },
    {
      "epoch": 0.13671700473594853,
      "grad_norm": 0.4919504225254059,
      "learning_rate": 2e-05,
      "loss": 0.7461,
      "step": 1530
    },
    {
      "epoch": 0.13761057993030112,
      "grad_norm": 0.37025532126426697,
      "learning_rate": 2e-05,
      "loss": 0.7226,
      "step": 1540
    },
    {
      "epoch": 0.13850415512465375,
      "grad_norm": 0.3523809611797333,
      "learning_rate": 2e-05,
      "loss": 0.7124,
      "step": 1550
    },
    {
      "epoch": 0.13939773031900635,
      "grad_norm": 0.4186336398124695,
      "learning_rate": 2e-05,
      "loss": 0.724,
      "step": 1560
    },
    {
      "epoch": 0.14029130551335894,
      "grad_norm": 0.39147084951400757,
      "learning_rate": 2e-05,
      "loss": 0.7444,
      "step": 1570
    },
    {
      "epoch": 0.14118488070771154,
      "grad_norm": 0.3777700364589691,
      "learning_rate": 2e-05,
      "loss": 0.7383,
      "step": 1580
    },
    {
      "epoch": 0.14207845590206417,
      "grad_norm": 0.41446128487586975,
      "learning_rate": 2e-05,
      "loss": 0.735,
      "step": 1590
    },
    {
      "epoch": 0.14297203109641676,
      "grad_norm": 0.38765832781791687,
      "learning_rate": 2e-05,
      "loss": 0.7147,
      "step": 1600
    },
    {
      "epoch": 0.14386560629076936,
      "grad_norm": 0.5072250962257385,
      "learning_rate": 2e-05,
      "loss": 0.7163,
      "step": 1610
    },
    {
      "epoch": 0.14475918148512198,
      "grad_norm": 0.3880365490913391,
      "learning_rate": 2e-05,
      "loss": 0.7141,
      "step": 1620
    },
    {
      "epoch": 0.14565275667947458,
      "grad_norm": 0.35895800590515137,
      "learning_rate": 2e-05,
      "loss": 0.7137,
      "step": 1630
    },
    {
      "epoch": 0.14654633187382718,
      "grad_norm": 0.35093289613723755,
      "learning_rate": 2e-05,
      "loss": 0.7359,
      "step": 1640
    },
    {
      "epoch": 0.14743990706817978,
      "grad_norm": 0.3536919951438904,
      "learning_rate": 2e-05,
      "loss": 0.739,
      "step": 1650
    },
    {
      "epoch": 0.1483334822625324,
      "grad_norm": 0.6556247472763062,
      "learning_rate": 2e-05,
      "loss": 0.7185,
      "step": 1660
    },
    {
      "epoch": 0.149227057456885,
      "grad_norm": 0.3924124538898468,
      "learning_rate": 2e-05,
      "loss": 0.747,
      "step": 1670
    },
    {
      "epoch": 0.1501206326512376,
      "grad_norm": 0.8066271543502808,
      "learning_rate": 2e-05,
      "loss": 0.7584,
      "step": 1680
    },
    {
      "epoch": 0.1510142078455902,
      "grad_norm": 0.4922969341278076,
      "learning_rate": 2e-05,
      "loss": 0.7044,
      "step": 1690
    },
    {
      "epoch": 0.15190778303994282,
      "grad_norm": 0.34611326456069946,
      "learning_rate": 2e-05,
      "loss": 0.7443,
      "step": 1700
    },
    {
      "epoch": 0.15280135823429541,
      "grad_norm": 0.3866153657436371,
      "learning_rate": 2e-05,
      "loss": 0.725,
      "step": 1710
    },
    {
      "epoch": 0.153694933428648,
      "grad_norm": 0.36212798953056335,
      "learning_rate": 2e-05,
      "loss": 0.7245,
      "step": 1720
    },
    {
      "epoch": 0.15458850862300064,
      "grad_norm": 0.3953538239002228,
      "learning_rate": 2e-05,
      "loss": 0.7347,
      "step": 1730
    },
    {
      "epoch": 0.15548208381735323,
      "grad_norm": 0.4102666676044464,
      "learning_rate": 2e-05,
      "loss": 0.736,
      "step": 1740
    },
    {
      "epoch": 0.15637565901170583,
      "grad_norm": 0.4497690498828888,
      "learning_rate": 2e-05,
      "loss": 0.7245,
      "step": 1750
    },
    {
      "epoch": 0.15726923420605843,
      "grad_norm": 0.4439946711063385,
      "learning_rate": 2e-05,
      "loss": 0.7347,
      "step": 1760
    },
    {
      "epoch": 0.15816280940041105,
      "grad_norm": 1.074998140335083,
      "learning_rate": 2e-05,
      "loss": 0.7432,
      "step": 1770
    },
    {
      "epoch": 0.15905638459476365,
      "grad_norm": 0.3673039972782135,
      "learning_rate": 2e-05,
      "loss": 0.7465,
      "step": 1780
    },
    {
      "epoch": 0.15994995978911625,
      "grad_norm": 0.3478325605392456,
      "learning_rate": 2e-05,
      "loss": 0.7225,
      "step": 1790
    },
    {
      "epoch": 0.16084353498346887,
      "grad_norm": 0.3802620470523834,
      "learning_rate": 2e-05,
      "loss": 0.6983,
      "step": 1800
    },
    {
      "epoch": 0.16173711017782147,
      "grad_norm": 0.38965746760368347,
      "learning_rate": 2e-05,
      "loss": 0.712,
      "step": 1810
    },
    {
      "epoch": 0.16263068537217407,
      "grad_norm": 0.3660908043384552,
      "learning_rate": 2e-05,
      "loss": 0.7198,
      "step": 1820
    },
    {
      "epoch": 0.16352426056652666,
      "grad_norm": 0.38565513491630554,
      "learning_rate": 2e-05,
      "loss": 0.7118,
      "step": 1830
    },
    {
      "epoch": 0.1644178357608793,
      "grad_norm": 0.49884864687919617,
      "learning_rate": 2e-05,
      "loss": 0.7378,
      "step": 1840
    },
    {
      "epoch": 0.16531141095523189,
      "grad_norm": 0.3706195652484894,
      "learning_rate": 2e-05,
      "loss": 0.7051,
      "step": 1850
    },
    {
      "epoch": 0.16620498614958448,
      "grad_norm": 0.3866501450538635,
      "learning_rate": 2e-05,
      "loss": 0.7321,
      "step": 1860
    },
    {
      "epoch": 0.16709856134393708,
      "grad_norm": 0.38853004574775696,
      "learning_rate": 2e-05,
      "loss": 0.7445,
      "step": 1870
    },
    {
      "epoch": 0.1679921365382897,
      "grad_norm": 0.42117682099342346,
      "learning_rate": 2e-05,
      "loss": 0.7241,
      "step": 1880
    },
    {
      "epoch": 0.1688857117326423,
      "grad_norm": 0.4492304027080536,
      "learning_rate": 2e-05,
      "loss": 0.7349,
      "step": 1890
    },
    {
      "epoch": 0.1697792869269949,
      "grad_norm": 0.4259665608406067,
      "learning_rate": 2e-05,
      "loss": 0.7128,
      "step": 1900
    },
    {
      "epoch": 0.17067286212134752,
      "grad_norm": 0.3557761609554291,
      "learning_rate": 2e-05,
      "loss": 0.6986,
      "step": 1910
    },
    {
      "epoch": 0.17156643731570012,
      "grad_norm": 0.35731881856918335,
      "learning_rate": 2e-05,
      "loss": 0.7351,
      "step": 1920
    },
    {
      "epoch": 0.17246001251005272,
      "grad_norm": 0.38917088508605957,
      "learning_rate": 2e-05,
      "loss": 0.721,
      "step": 1930
    },
    {
      "epoch": 0.17335358770440532,
      "grad_norm": 0.39296889305114746,
      "learning_rate": 2e-05,
      "loss": 0.7457,
      "step": 1940
    },
    {
      "epoch": 0.17424716289875794,
      "grad_norm": 0.3701498508453369,
      "learning_rate": 2e-05,
      "loss": 0.7699,
      "step": 1950
    },
    {
      "epoch": 0.17514073809311054,
      "grad_norm": 0.3800022006034851,
      "learning_rate": 2e-05,
      "loss": 0.7222,
      "step": 1960
    },
    {
      "epoch": 0.17603431328746313,
      "grad_norm": 0.3638252913951874,
      "learning_rate": 2e-05,
      "loss": 0.6871,
      "step": 1970
    },
    {
      "epoch": 0.17692788848181573,
      "grad_norm": 0.38935279846191406,
      "learning_rate": 2e-05,
      "loss": 0.6952,
      "step": 1980
    },
    {
      "epoch": 0.17782146367616836,
      "grad_norm": 0.40048500895500183,
      "learning_rate": 2e-05,
      "loss": 0.697,
      "step": 1990
    },
    {
      "epoch": 0.17871503887052095,
      "grad_norm": 0.41115978360176086,
      "learning_rate": 2e-05,
      "loss": 0.7409,
      "step": 2000
    },
    {
      "epoch": 0.17960861406487355,
      "grad_norm": 0.39649268984794617,
      "learning_rate": 2e-05,
      "loss": 0.7172,
      "step": 2010
    },
    {
      "epoch": 0.18050218925922618,
      "grad_norm": 0.5061083436012268,
      "learning_rate": 2e-05,
      "loss": 0.7264,
      "step": 2020
    },
    {
      "epoch": 0.18139576445357877,
      "grad_norm": 0.38063013553619385,
      "learning_rate": 2e-05,
      "loss": 0.7434,
      "step": 2030
    },
    {
      "epoch": 0.18228933964793137,
      "grad_norm": 0.4168822169303894,
      "learning_rate": 2e-05,
      "loss": 0.7063,
      "step": 2040
    },
    {
      "epoch": 0.18318291484228397,
      "grad_norm": 0.36266323924064636,
      "learning_rate": 2e-05,
      "loss": 0.7401,
      "step": 2050
    },
    {
      "epoch": 0.1840764900366366,
      "grad_norm": 0.37227675318717957,
      "learning_rate": 2e-05,
      "loss": 0.7144,
      "step": 2060
    },
    {
      "epoch": 0.1849700652309892,
      "grad_norm": 0.39245229959487915,
      "learning_rate": 2e-05,
      "loss": 0.7084,
      "step": 2070
    },
    {
      "epoch": 0.1858636404253418,
      "grad_norm": 0.37118712067604065,
      "learning_rate": 2e-05,
      "loss": 0.7217,
      "step": 2080
    },
    {
      "epoch": 0.1867572156196944,
      "grad_norm": 0.3615480661392212,
      "learning_rate": 2e-05,
      "loss": 0.7143,
      "step": 2090
    },
    {
      "epoch": 0.187650790814047,
      "grad_norm": 0.4019179046154022,
      "learning_rate": 2e-05,
      "loss": 0.7449,
      "step": 2100
    },
    {
      "epoch": 0.1885443660083996,
      "grad_norm": 0.38081419467926025,
      "learning_rate": 2e-05,
      "loss": 0.724,
      "step": 2110
    },
    {
      "epoch": 0.1894379412027522,
      "grad_norm": 0.39533179998397827,
      "learning_rate": 2e-05,
      "loss": 0.7065,
      "step": 2120
    },
    {
      "epoch": 0.19033151639710483,
      "grad_norm": 0.4144379794597626,
      "learning_rate": 2e-05,
      "loss": 0.7177,
      "step": 2130
    },
    {
      "epoch": 0.19122509159145742,
      "grad_norm": 0.38099774718284607,
      "learning_rate": 2e-05,
      "loss": 0.7035,
      "step": 2140
    },
    {
      "epoch": 0.19211866678581002,
      "grad_norm": 0.36082664132118225,
      "learning_rate": 2e-05,
      "loss": 0.729,
      "step": 2150
    },
    {
      "epoch": 0.19301224198016262,
      "grad_norm": 0.38504886627197266,
      "learning_rate": 2e-05,
      "loss": 0.7049,
      "step": 2160
    },
    {
      "epoch": 0.19390581717451524,
      "grad_norm": 0.3863053619861603,
      "learning_rate": 2e-05,
      "loss": 0.6846,
      "step": 2170
    },
    {
      "epoch": 0.19479939236886784,
      "grad_norm": 0.3692690432071686,
      "learning_rate": 2e-05,
      "loss": 0.7298,
      "step": 2180
    },
    {
      "epoch": 0.19569296756322044,
      "grad_norm": 0.3638010621070862,
      "learning_rate": 2e-05,
      "loss": 0.7256,
      "step": 2190
    },
    {
      "epoch": 0.19658654275757306,
      "grad_norm": 0.3824363648891449,
      "learning_rate": 2e-05,
      "loss": 0.7376,
      "step": 2200
    },
    {
      "epoch": 0.19748011795192566,
      "grad_norm": 0.35318830609321594,
      "learning_rate": 2e-05,
      "loss": 0.736,
      "step": 2210
    },
    {
      "epoch": 0.19837369314627826,
      "grad_norm": 0.35539188981056213,
      "learning_rate": 2e-05,
      "loss": 0.7027,
      "step": 2220
    },
    {
      "epoch": 0.19926726834063085,
      "grad_norm": 0.414533406496048,
      "learning_rate": 2e-05,
      "loss": 0.6909,
      "step": 2230
    },
    {
      "epoch": 0.20016084353498348,
      "grad_norm": 0.39531072974205017,
      "learning_rate": 2e-05,
      "loss": 0.7301,
      "step": 2240
    },
    {
      "epoch": 0.20105441872933608,
      "grad_norm": 0.3679538071155548,
      "learning_rate": 2e-05,
      "loss": 0.7203,
      "step": 2250
    },
    {
      "epoch": 0.20194799392368867,
      "grad_norm": 0.37164831161499023,
      "learning_rate": 2e-05,
      "loss": 0.725,
      "step": 2260
    },
    {
      "epoch": 0.20284156911804127,
      "grad_norm": 0.3633818030357361,
      "learning_rate": 2e-05,
      "loss": 0.7192,
      "step": 2270
    },
    {
      "epoch": 0.2037351443123939,
      "grad_norm": 0.3682442307472229,
      "learning_rate": 2e-05,
      "loss": 0.7099,
      "step": 2280
    },
    {
      "epoch": 0.2046287195067465,
      "grad_norm": 0.34997084736824036,
      "learning_rate": 2e-05,
      "loss": 0.7543,
      "step": 2290
    },
    {
      "epoch": 0.2055222947010991,
      "grad_norm": 0.403976172208786,
      "learning_rate": 2e-05,
      "loss": 0.7511,
      "step": 2300
    },
    {
      "epoch": 0.20641586989545171,
      "grad_norm": 0.41093239188194275,
      "learning_rate": 2e-05,
      "loss": 0.6794,
      "step": 2310
    },
    {
      "epoch": 0.2073094450898043,
      "grad_norm": 0.38257455825805664,
      "learning_rate": 2e-05,
      "loss": 0.7263,
      "step": 2320
    },
    {
      "epoch": 0.2082030202841569,
      "grad_norm": 0.3526884615421295,
      "learning_rate": 2e-05,
      "loss": 0.7185,
      "step": 2330
    },
    {
      "epoch": 0.2090965954785095,
      "grad_norm": 0.3658798336982727,
      "learning_rate": 2e-05,
      "loss": 0.7128,
      "step": 2340
    },
    {
      "epoch": 0.20999017067286213,
      "grad_norm": 0.3977317214012146,
      "learning_rate": 2e-05,
      "loss": 0.6795,
      "step": 2350
    },
    {
      "epoch": 0.21088374586721473,
      "grad_norm": 0.386541485786438,
      "learning_rate": 2e-05,
      "loss": 0.7196,
      "step": 2360
    },
    {
      "epoch": 0.21177732106156733,
      "grad_norm": 0.39826154708862305,
      "learning_rate": 2e-05,
      "loss": 0.6952,
      "step": 2370
    },
    {
      "epoch": 0.21267089625591992,
      "grad_norm": 0.4097355902194977,
      "learning_rate": 2e-05,
      "loss": 0.7117,
      "step": 2380
    },
    {
      "epoch": 0.21356447145027255,
      "grad_norm": 0.3464928865432739,
      "learning_rate": 2e-05,
      "loss": 0.7126,
      "step": 2390
    },
    {
      "epoch": 0.21445804664462514,
      "grad_norm": 0.397562712430954,
      "learning_rate": 2e-05,
      "loss": 0.7197,
      "step": 2400
    },
    {
      "epoch": 0.21535162183897774,
      "grad_norm": 0.35849079489707947,
      "learning_rate": 2e-05,
      "loss": 0.7525,
      "step": 2410
    },
    {
      "epoch": 0.21624519703333037,
      "grad_norm": 0.5133382081985474,
      "learning_rate": 2e-05,
      "loss": 0.7267,
      "step": 2420
    },
    {
      "epoch": 0.21713877222768296,
      "grad_norm": 0.37571123242378235,
      "learning_rate": 2e-05,
      "loss": 0.7223,
      "step": 2430
    },
    {
      "epoch": 0.21803234742203556,
      "grad_norm": 0.3847047686576843,
      "learning_rate": 2e-05,
      "loss": 0.7139,
      "step": 2440
    },
    {
      "epoch": 0.21892592261638816,
      "grad_norm": 0.3535878360271454,
      "learning_rate": 2e-05,
      "loss": 0.6956,
      "step": 2450
    },
    {
      "epoch": 0.21981949781074078,
      "grad_norm": 0.3725936710834503,
      "learning_rate": 2e-05,
      "loss": 0.7107,
      "step": 2460
    },
    {
      "epoch": 0.22071307300509338,
      "grad_norm": 0.4022907316684723,
      "learning_rate": 2e-05,
      "loss": 0.7524,
      "step": 2470
    },
    {
      "epoch": 0.22160664819944598,
      "grad_norm": 0.3813494145870209,
      "learning_rate": 2e-05,
      "loss": 0.7229,
      "step": 2480
    },
    {
      "epoch": 0.2225002233937986,
      "grad_norm": 0.38350072503089905,
      "learning_rate": 2e-05,
      "loss": 0.6846,
      "step": 2490
    },
    {
      "epoch": 0.2233937985881512,
      "grad_norm": 0.3981742858886719,
      "learning_rate": 2e-05,
      "loss": 0.7362,
      "step": 2500
    },
    {
      "epoch": 0.2242873737825038,
      "grad_norm": 0.3835667371749878,
      "learning_rate": 2e-05,
      "loss": 0.7107,
      "step": 2510
    },
    {
      "epoch": 0.2251809489768564,
      "grad_norm": 0.372981458902359,
      "learning_rate": 2e-05,
      "loss": 0.7333,
      "step": 2520
    },
    {
      "epoch": 0.22607452417120902,
      "grad_norm": 0.35782164335250854,
      "learning_rate": 2e-05,
      "loss": 0.6753,
      "step": 2530
    },
    {
      "epoch": 0.22696809936556162,
      "grad_norm": 0.36539599299430847,
      "learning_rate": 2e-05,
      "loss": 0.6941,
      "step": 2540
    },
    {
      "epoch": 0.2278616745599142,
      "grad_norm": 0.3650493919849396,
      "learning_rate": 2e-05,
      "loss": 0.7003,
      "step": 2550
    },
    {
      "epoch": 0.2287552497542668,
      "grad_norm": 0.3873494863510132,
      "learning_rate": 2e-05,
      "loss": 0.7172,
      "step": 2560
    },
    {
      "epoch": 0.22964882494861943,
      "grad_norm": 0.3694900870323181,
      "learning_rate": 2e-05,
      "loss": 0.7238,
      "step": 2570
    },
    {
      "epoch": 0.23054240014297203,
      "grad_norm": 0.3531987965106964,
      "learning_rate": 2e-05,
      "loss": 0.7374,
      "step": 2580
    },
    {
      "epoch": 0.23143597533732463,
      "grad_norm": 0.3426517844200134,
      "learning_rate": 2e-05,
      "loss": 0.7285,
      "step": 2590
    },
    {
      "epoch": 0.23232955053167725,
      "grad_norm": 0.3928007185459137,
      "learning_rate": 2e-05,
      "loss": 0.7108,
      "step": 2600
    },
    {
      "epoch": 0.23322312572602985,
      "grad_norm": 0.37236160039901733,
      "learning_rate": 2e-05,
      "loss": 0.7158,
      "step": 2610
    },
    {
      "epoch": 0.23411670092038245,
      "grad_norm": 0.38145866990089417,
      "learning_rate": 2e-05,
      "loss": 0.6846,
      "step": 2620
    },
    {
      "epoch": 0.23501027611473504,
      "grad_norm": 0.3722963333129883,
      "learning_rate": 2e-05,
      "loss": 0.7169,
      "step": 2630
    },
    {
      "epoch": 0.23590385130908767,
      "grad_norm": 0.42692744731903076,
      "learning_rate": 2e-05,
      "loss": 0.7264,
      "step": 2640
    },
    {
      "epoch": 0.23679742650344027,
      "grad_norm": 0.37643131613731384,
      "learning_rate": 2e-05,
      "loss": 0.7307,
      "step": 2650
    },
    {
      "epoch": 0.23769100169779286,
      "grad_norm": 0.3665594756603241,
      "learning_rate": 2e-05,
      "loss": 0.7002,
      "step": 2660
    },
    {
      "epoch": 0.23858457689214546,
      "grad_norm": 0.4624324142932892,
      "learning_rate": 2e-05,
      "loss": 0.7116,
      "step": 2670
    },
    {
      "epoch": 0.2394781520864981,
      "grad_norm": 0.3787866234779358,
      "learning_rate": 2e-05,
      "loss": 0.7084,
      "step": 2680
    },
    {
      "epoch": 0.24037172728085068,
      "grad_norm": 0.3793836236000061,
      "learning_rate": 2e-05,
      "loss": 0.7268,
      "step": 2690
    },
    {
      "epoch": 0.24126530247520328,
      "grad_norm": 0.3420543670654297,
      "learning_rate": 2e-05,
      "loss": 0.7109,
      "step": 2700
    },
    {
      "epoch": 0.2421588776695559,
      "grad_norm": 0.3823523223400116,
      "learning_rate": 2e-05,
      "loss": 0.7257,
      "step": 2710
    },
    {
      "epoch": 0.2430524528639085,
      "grad_norm": 0.35775208473205566,
      "learning_rate": 2e-05,
      "loss": 0.6907,
      "step": 2720
    },
    {
      "epoch": 0.2439460280582611,
      "grad_norm": 0.36065220832824707,
      "learning_rate": 2e-05,
      "loss": 0.7154,
      "step": 2730
    },
    {
      "epoch": 0.2448396032526137,
      "grad_norm": 0.3732667565345764,
      "learning_rate": 2e-05,
      "loss": 0.7,
      "step": 2740
    },
    {
      "epoch": 0.24573317844696632,
      "grad_norm": 0.34359318017959595,
      "learning_rate": 2e-05,
      "loss": 0.7244,
      "step": 2750
    },
    {
      "epoch": 0.24662675364131892,
      "grad_norm": 0.35983023047447205,
      "learning_rate": 2e-05,
      "loss": 0.6753,
      "step": 2760
    },
    {
      "epoch": 0.24752032883567152,
      "grad_norm": 0.37808793783187866,
      "learning_rate": 2e-05,
      "loss": 0.7379,
      "step": 2770
    },
    {
      "epoch": 0.2484139040300241,
      "grad_norm": 0.38024666905403137,
      "learning_rate": 2e-05,
      "loss": 0.7081,
      "step": 2780
    },
    {
      "epoch": 0.24930747922437674,
      "grad_norm": 0.3341760039329529,
      "learning_rate": 2e-05,
      "loss": 0.7211,
      "step": 2790
    },
    {
      "epoch": 0.25020105441872936,
      "grad_norm": 0.363836407661438,
      "learning_rate": 2e-05,
      "loss": 0.6929,
      "step": 2800
    },
    {
      "epoch": 0.25109462961308193,
      "grad_norm": 0.36290687322616577,
      "learning_rate": 2e-05,
      "loss": 0.7013,
      "step": 2810
    },
    {
      "epoch": 0.25198820480743456,
      "grad_norm": 0.3314644396305084,
      "learning_rate": 2e-05,
      "loss": 0.7216,
      "step": 2820
    },
    {
      "epoch": 0.2528817800017871,
      "grad_norm": 0.3978954255580902,
      "learning_rate": 2e-05,
      "loss": 0.7301,
      "step": 2830
    },
    {
      "epoch": 0.25377535519613975,
      "grad_norm": 0.37203192710876465,
      "learning_rate": 2e-05,
      "loss": 0.7025,
      "step": 2840
    },
    {
      "epoch": 0.2546689303904924,
      "grad_norm": 0.3674238622188568,
      "learning_rate": 2e-05,
      "loss": 0.6771,
      "step": 2850
    },
    {
      "epoch": 0.25556250558484495,
      "grad_norm": 0.398285448551178,
      "learning_rate": 2e-05,
      "loss": 0.7265,
      "step": 2860
    },
    {
      "epoch": 0.25645608077919757,
      "grad_norm": 0.4424970746040344,
      "learning_rate": 2e-05,
      "loss": 0.7293,
      "step": 2870
    },
    {
      "epoch": 0.2573496559735502,
      "grad_norm": 0.34791824221611023,
      "learning_rate": 2e-05,
      "loss": 0.7064,
      "step": 2880
    },
    {
      "epoch": 0.25824323116790276,
      "grad_norm": 0.3702598214149475,
      "learning_rate": 2e-05,
      "loss": 0.7207,
      "step": 2890
    },
    {
      "epoch": 0.2591368063622554,
      "grad_norm": 0.36925578117370605,
      "learning_rate": 2e-05,
      "loss": 0.6949,
      "step": 2900
    },
    {
      "epoch": 0.260030381556608,
      "grad_norm": 0.3745940923690796,
      "learning_rate": 2e-05,
      "loss": 0.7172,
      "step": 2910
    },
    {
      "epoch": 0.2609239567509606,
      "grad_norm": 0.3436862528324127,
      "learning_rate": 2e-05,
      "loss": 0.6852,
      "step": 2920
    },
    {
      "epoch": 0.2618175319453132,
      "grad_norm": 0.40740537643432617,
      "learning_rate": 2e-05,
      "loss": 0.6863,
      "step": 2930
    },
    {
      "epoch": 0.2627111071396658,
      "grad_norm": 0.398599773645401,
      "learning_rate": 2e-05,
      "loss": 0.7181,
      "step": 2940
    },
    {
      "epoch": 0.2636046823340184,
      "grad_norm": 0.37414512038230896,
      "learning_rate": 2e-05,
      "loss": 0.6983,
      "step": 2950
    },
    {
      "epoch": 0.26449825752837103,
      "grad_norm": 0.39800015091896057,
      "learning_rate": 2e-05,
      "loss": 0.7525,
      "step": 2960
    },
    {
      "epoch": 0.2653918327227236,
      "grad_norm": 0.37266287207603455,
      "learning_rate": 2e-05,
      "loss": 0.715,
      "step": 2970
    },
    {
      "epoch": 0.2662854079170762,
      "grad_norm": 0.35028383135795593,
      "learning_rate": 2e-05,
      "loss": 0.7252,
      "step": 2980
    },
    {
      "epoch": 0.26717898311142885,
      "grad_norm": 0.36741185188293457,
      "learning_rate": 2e-05,
      "loss": 0.7095,
      "step": 2990
    },
    {
      "epoch": 0.2680725583057814,
      "grad_norm": 0.3469706177711487,
      "learning_rate": 2e-05,
      "loss": 0.7302,
      "step": 3000
    },
    {
      "epoch": 0.26896613350013404,
      "grad_norm": 0.38094988465309143,
      "learning_rate": 2e-05,
      "loss": 0.7099,
      "step": 3010
    },
    {
      "epoch": 0.26985970869448667,
      "grad_norm": 0.34936171770095825,
      "learning_rate": 2e-05,
      "loss": 0.6844,
      "step": 3020
    },
    {
      "epoch": 0.27075328388883924,
      "grad_norm": 0.3987492620944977,
      "learning_rate": 2e-05,
      "loss": 0.6901,
      "step": 3030
    },
    {
      "epoch": 0.27164685908319186,
      "grad_norm": 0.39074254035949707,
      "learning_rate": 2e-05,
      "loss": 0.7103,
      "step": 3040
    },
    {
      "epoch": 0.27254043427754443,
      "grad_norm": 0.4020327925682068,
      "learning_rate": 2e-05,
      "loss": 0.7226,
      "step": 3050
    },
    {
      "epoch": 0.27343400947189705,
      "grad_norm": 0.6894024014472961,
      "learning_rate": 2e-05,
      "loss": 0.7021,
      "step": 3060
    },
    {
      "epoch": 0.2743275846662497,
      "grad_norm": 0.41694730520248413,
      "learning_rate": 2e-05,
      "loss": 0.7251,
      "step": 3070
    },
    {
      "epoch": 0.27522115986060225,
      "grad_norm": 0.3496346175670624,
      "learning_rate": 2e-05,
      "loss": 0.702,
      "step": 3080
    },
    {
      "epoch": 0.2761147350549549,
      "grad_norm": 0.35392478108406067,
      "learning_rate": 2e-05,
      "loss": 0.7099,
      "step": 3090
    },
    {
      "epoch": 0.2770083102493075,
      "grad_norm": 0.3742842674255371,
      "learning_rate": 2e-05,
      "loss": 0.704,
      "step": 3100
    },
    {
      "epoch": 0.27790188544366007,
      "grad_norm": 0.36130213737487793,
      "learning_rate": 2e-05,
      "loss": 0.7179,
      "step": 3110
    },
    {
      "epoch": 0.2787954606380127,
      "grad_norm": 0.36318132281303406,
      "learning_rate": 2e-05,
      "loss": 0.7,
      "step": 3120
    },
    {
      "epoch": 0.2796890358323653,
      "grad_norm": 0.40958553552627563,
      "learning_rate": 2e-05,
      "loss": 0.7042,
      "step": 3130
    },
    {
      "epoch": 0.2805826110267179,
      "grad_norm": 0.3563935458660126,
      "learning_rate": 2e-05,
      "loss": 0.7358,
      "step": 3140
    },
    {
      "epoch": 0.2814761862210705,
      "grad_norm": 0.3450090289115906,
      "learning_rate": 2e-05,
      "loss": 0.6818,
      "step": 3150
    },
    {
      "epoch": 0.2823697614154231,
      "grad_norm": 0.40280184149742126,
      "learning_rate": 2e-05,
      "loss": 0.675,
      "step": 3160
    },
    {
      "epoch": 0.2832633366097757,
      "grad_norm": 0.37556108832359314,
      "learning_rate": 2e-05,
      "loss": 0.7182,
      "step": 3170
    },
    {
      "epoch": 0.28415691180412833,
      "grad_norm": 0.41327449679374695,
      "learning_rate": 2e-05,
      "loss": 0.6926,
      "step": 3180
    },
    {
      "epoch": 0.2850504869984809,
      "grad_norm": 0.33954551815986633,
      "learning_rate": 2e-05,
      "loss": 0.7161,
      "step": 3190
    },
    {
      "epoch": 0.2859440621928335,
      "grad_norm": 0.3450734317302704,
      "learning_rate": 2e-05,
      "loss": 0.7083,
      "step": 3200
    },
    {
      "epoch": 0.28683763738718615,
      "grad_norm": 0.3419007360935211,
      "learning_rate": 2e-05,
      "loss": 0.7117,
      "step": 3210
    },
    {
      "epoch": 0.2877312125815387,
      "grad_norm": 0.4342067837715149,
      "learning_rate": 2e-05,
      "loss": 0.6925,
      "step": 3220
    },
    {
      "epoch": 0.28862478777589134,
      "grad_norm": 0.39862027764320374,
      "learning_rate": 2e-05,
      "loss": 0.7058,
      "step": 3230
    },
    {
      "epoch": 0.28951836297024397,
      "grad_norm": 0.37627285718917847,
      "learning_rate": 2e-05,
      "loss": 0.7113,
      "step": 3240
    },
    {
      "epoch": 0.29041193816459654,
      "grad_norm": 0.36114147305488586,
      "learning_rate": 2e-05,
      "loss": 0.7084,
      "step": 3250
    },
    {
      "epoch": 0.29130551335894916,
      "grad_norm": 0.3913547098636627,
      "learning_rate": 2e-05,
      "loss": 0.7155,
      "step": 3260
    },
    {
      "epoch": 0.29219908855330173,
      "grad_norm": 0.48569539189338684,
      "learning_rate": 2e-05,
      "loss": 0.6851,
      "step": 3270
    },
    {
      "epoch": 0.29309266374765436,
      "grad_norm": 0.35713067650794983,
      "learning_rate": 2e-05,
      "loss": 0.6896,
      "step": 3280
    },
    {
      "epoch": 0.293986238942007,
      "grad_norm": 0.3882346451282501,
      "learning_rate": 2e-05,
      "loss": 0.7099,
      "step": 3290
    },
    {
      "epoch": 0.29487981413635955,
      "grad_norm": 0.4093131124973297,
      "learning_rate": 2e-05,
      "loss": 0.7114,
      "step": 3300
    },
    {
      "epoch": 0.2957733893307122,
      "grad_norm": 0.3757636845111847,
      "learning_rate": 2e-05,
      "loss": 0.7258,
      "step": 3310
    },
    {
      "epoch": 0.2966669645250648,
      "grad_norm": 0.4188380837440491,
      "learning_rate": 2e-05,
      "loss": 0.7065,
      "step": 3320
    },
    {
      "epoch": 0.29756053971941737,
      "grad_norm": 0.4249805510044098,
      "learning_rate": 2e-05,
      "loss": 0.6834,
      "step": 3330
    },
    {
      "epoch": 0.29845411491377,
      "grad_norm": 0.37131360173225403,
      "learning_rate": 2e-05,
      "loss": 0.6813,
      "step": 3340
    },
    {
      "epoch": 0.2993476901081226,
      "grad_norm": 0.4095926582813263,
      "learning_rate": 2e-05,
      "loss": 0.6989,
      "step": 3350
    },
    {
      "epoch": 0.3002412653024752,
      "grad_norm": 0.40671223402023315,
      "learning_rate": 2e-05,
      "loss": 0.6926,
      "step": 3360
    },
    {
      "epoch": 0.3011348404968278,
      "grad_norm": 0.3677366077899933,
      "learning_rate": 2e-05,
      "loss": 0.6928,
      "step": 3370
    },
    {
      "epoch": 0.3020284156911804,
      "grad_norm": 0.38548967242240906,
      "learning_rate": 2e-05,
      "loss": 0.7158,
      "step": 3380
    },
    {
      "epoch": 0.302921990885533,
      "grad_norm": 0.38187336921691895,
      "learning_rate": 2e-05,
      "loss": 0.7225,
      "step": 3390
    },
    {
      "epoch": 0.30381556607988564,
      "grad_norm": 0.3772178590297699,
      "learning_rate": 2e-05,
      "loss": 0.6714,
      "step": 3400
    },
    {
      "epoch": 0.3047091412742382,
      "grad_norm": 0.36621707677841187,
      "learning_rate": 2e-05,
      "loss": 0.7108,
      "step": 3410
    },
    {
      "epoch": 0.30560271646859083,
      "grad_norm": 0.3519044816493988,
      "learning_rate": 2e-05,
      "loss": 0.6995,
      "step": 3420
    },
    {
      "epoch": 0.30649629166294345,
      "grad_norm": 0.362652987241745,
      "learning_rate": 2e-05,
      "loss": 0.7082,
      "step": 3430
    },
    {
      "epoch": 0.307389866857296,
      "grad_norm": 0.3681955635547638,
      "learning_rate": 2e-05,
      "loss": 0.7242,
      "step": 3440
    },
    {
      "epoch": 0.30828344205164865,
      "grad_norm": 0.36930564045906067,
      "learning_rate": 2e-05,
      "loss": 0.731,
      "step": 3450
    },
    {
      "epoch": 0.3091770172460013,
      "grad_norm": 0.36028867959976196,
      "learning_rate": 2e-05,
      "loss": 0.7353,
      "step": 3460
    },
    {
      "epoch": 0.31007059244035384,
      "grad_norm": 0.3412564992904663,
      "learning_rate": 2e-05,
      "loss": 0.6911,
      "step": 3470
    },
    {
      "epoch": 0.31096416763470647,
      "grad_norm": 0.3616200387477875,
      "learning_rate": 2e-05,
      "loss": 0.6709,
      "step": 3480
    },
    {
      "epoch": 0.3118577428290591,
      "grad_norm": 0.3621264100074768,
      "learning_rate": 2e-05,
      "loss": 0.6842,
      "step": 3490
    },
    {
      "epoch": 0.31275131802341166,
      "grad_norm": 0.38668254017829895,
      "learning_rate": 2e-05,
      "loss": 0.7577,
      "step": 3500
    },
    {
      "epoch": 0.3136448932177643,
      "grad_norm": 0.40664446353912354,
      "learning_rate": 2e-05,
      "loss": 0.6919,
      "step": 3510
    },
    {
      "epoch": 0.31453846841211686,
      "grad_norm": 0.4402414858341217,
      "learning_rate": 2e-05,
      "loss": 0.7077,
      "step": 3520
    },
    {
      "epoch": 0.3154320436064695,
      "grad_norm": 0.3952922821044922,
      "learning_rate": 2e-05,
      "loss": 0.6773,
      "step": 3530
    },
    {
      "epoch": 0.3163256188008221,
      "grad_norm": 0.4013434946537018,
      "learning_rate": 2e-05,
      "loss": 0.6899,
      "step": 3540
    },
    {
      "epoch": 0.3172191939951747,
      "grad_norm": 0.37860897183418274,
      "learning_rate": 2e-05,
      "loss": 0.695,
      "step": 3550
    },
    {
      "epoch": 0.3181127691895273,
      "grad_norm": 0.397117555141449,
      "learning_rate": 2e-05,
      "loss": 0.6925,
      "step": 3560
    },
    {
      "epoch": 0.3190063443838799,
      "grad_norm": 0.46799492835998535,
      "learning_rate": 2e-05,
      "loss": 0.7116,
      "step": 3570
    },
    {
      "epoch": 0.3198999195782325,
      "grad_norm": 0.34963753819465637,
      "learning_rate": 2e-05,
      "loss": 0.7115,
      "step": 3580
    },
    {
      "epoch": 0.3207934947725851,
      "grad_norm": 0.411518931388855,
      "learning_rate": 2e-05,
      "loss": 0.6943,
      "step": 3590
    },
    {
      "epoch": 0.32168706996693774,
      "grad_norm": 0.36255308985710144,
      "learning_rate": 2e-05,
      "loss": 0.7075,
      "step": 3600
    },
    {
      "epoch": 0.3225806451612903,
      "grad_norm": 0.3666975796222687,
      "learning_rate": 2e-05,
      "loss": 0.7209,
      "step": 3610
    },
    {
      "epoch": 0.32347422035564294,
      "grad_norm": 0.3676561713218689,
      "learning_rate": 2e-05,
      "loss": 0.6981,
      "step": 3620
    },
    {
      "epoch": 0.3243677955499955,
      "grad_norm": 0.417124480009079,
      "learning_rate": 2e-05,
      "loss": 0.7004,
      "step": 3630
    },
    {
      "epoch": 0.32526137074434813,
      "grad_norm": 0.3794247806072235,
      "learning_rate": 2e-05,
      "loss": 0.697,
      "step": 3640
    },
    {
      "epoch": 0.32615494593870076,
      "grad_norm": 0.40678223967552185,
      "learning_rate": 2e-05,
      "loss": 0.7257,
      "step": 3650
    },
    {
      "epoch": 0.3270485211330533,
      "grad_norm": 0.37268680334091187,
      "learning_rate": 2e-05,
      "loss": 0.6853,
      "step": 3660
    },
    {
      "epoch": 0.32794209632740595,
      "grad_norm": 0.38034191727638245,
      "learning_rate": 2e-05,
      "loss": 0.6743,
      "step": 3670
    },
    {
      "epoch": 0.3288356715217586,
      "grad_norm": 0.35265952348709106,
      "learning_rate": 2e-05,
      "loss": 0.7162,
      "step": 3680
    },
    {
      "epoch": 0.32972924671611115,
      "grad_norm": 0.36596548557281494,
      "learning_rate": 2e-05,
      "loss": 0.6653,
      "step": 3690
    },
    {
      "epoch": 0.33062282191046377,
      "grad_norm": 0.37113913893699646,
      "learning_rate": 2e-05,
      "loss": 0.6983,
      "step": 3700
    },
    {
      "epoch": 0.3315163971048164,
      "grad_norm": 0.4150978624820709,
      "learning_rate": 2e-05,
      "loss": 0.7174,
      "step": 3710
    },
    {
      "epoch": 0.33240997229916897,
      "grad_norm": 0.34214451909065247,
      "learning_rate": 2e-05,
      "loss": 0.7131,
      "step": 3720
    },
    {
      "epoch": 0.3333035474935216,
      "grad_norm": 0.6437499523162842,
      "learning_rate": 2e-05,
      "loss": 0.694,
      "step": 3730
    },
    {
      "epoch": 0.33419712268787416,
      "grad_norm": 0.33215585350990295,
      "learning_rate": 2e-05,
      "loss": 0.7026,
      "step": 3740
    },
    {
      "epoch": 0.3350906978822268,
      "grad_norm": 0.36878666281700134,
      "learning_rate": 2e-05,
      "loss": 0.714,
      "step": 3750
    },
    {
      "epoch": 0.3359842730765794,
      "grad_norm": 0.36607626080513,
      "learning_rate": 2e-05,
      "loss": 0.7004,
      "step": 3760
    },
    {
      "epoch": 0.336877848270932,
      "grad_norm": 0.41994553804397583,
      "learning_rate": 2e-05,
      "loss": 0.7132,
      "step": 3770
    },
    {
      "epoch": 0.3377714234652846,
      "grad_norm": 0.3406926691532135,
      "learning_rate": 2e-05,
      "loss": 0.669,
      "step": 3780
    },
    {
      "epoch": 0.33866499865963723,
      "grad_norm": 0.3424694538116455,
      "learning_rate": 2e-05,
      "loss": 0.7139,
      "step": 3790
    },
    {
      "epoch": 0.3395585738539898,
      "grad_norm": 0.3926262855529785,
      "learning_rate": 2e-05,
      "loss": 0.6655,
      "step": 3800
    },
    {
      "epoch": 0.3404521490483424,
      "grad_norm": 0.40683773159980774,
      "learning_rate": 2e-05,
      "loss": 0.7186,
      "step": 3810
    },
    {
      "epoch": 0.34134572424269505,
      "grad_norm": 0.3801291286945343,
      "learning_rate": 2e-05,
      "loss": 0.6809,
      "step": 3820
    },
    {
      "epoch": 0.3422392994370476,
      "grad_norm": 0.3531859517097473,
      "learning_rate": 2e-05,
      "loss": 0.684,
      "step": 3830
    },
    {
      "epoch": 0.34313287463140024,
      "grad_norm": 0.4508625268936157,
      "learning_rate": 2e-05,
      "loss": 0.6952,
      "step": 3840
    },
    {
      "epoch": 0.3440264498257528,
      "grad_norm": 0.3657115399837494,
      "learning_rate": 2e-05,
      "loss": 0.6673,
      "step": 3850
    },
    {
      "epoch": 0.34492002502010544,
      "grad_norm": 0.3592742085456848,
      "learning_rate": 2e-05,
      "loss": 0.6633,
      "step": 3860
    },
    {
      "epoch": 0.34581360021445806,
      "grad_norm": 0.36773285269737244,
      "learning_rate": 2e-05,
      "loss": 0.6801,
      "step": 3870
    },
    {
      "epoch": 0.34670717540881063,
      "grad_norm": 0.37082305550575256,
      "learning_rate": 2e-05,
      "loss": 0.6841,
      "step": 3880
    },
    {
      "epoch": 0.34760075060316326,
      "grad_norm": 0.3258555233478546,
      "learning_rate": 2e-05,
      "loss": 0.693,
      "step": 3890
    },
    {
      "epoch": 0.3484943257975159,
      "grad_norm": 0.4185699224472046,
      "learning_rate": 2e-05,
      "loss": 0.7062,
      "step": 3900
    },
    {
      "epoch": 0.34938790099186845,
      "grad_norm": 0.41331052780151367,
      "learning_rate": 2e-05,
      "loss": 0.6871,
      "step": 3910
    },
    {
      "epoch": 0.3502814761862211,
      "grad_norm": 0.36346831917762756,
      "learning_rate": 2e-05,
      "loss": 0.6869,
      "step": 3920
    },
    {
      "epoch": 0.3511750513805737,
      "grad_norm": 0.3878091871738434,
      "learning_rate": 2e-05,
      "loss": 0.6991,
      "step": 3930
    },
    {
      "epoch": 0.35206862657492627,
      "grad_norm": 0.3615358769893646,
      "learning_rate": 2e-05,
      "loss": 0.6847,
      "step": 3940
    },
    {
      "epoch": 0.3529622017692789,
      "grad_norm": 0.3688424229621887,
      "learning_rate": 2e-05,
      "loss": 0.6893,
      "step": 3950
    },
    {
      "epoch": 0.35385577696363146,
      "grad_norm": 0.40755635499954224,
      "learning_rate": 2e-05,
      "loss": 0.6959,
      "step": 3960
    },
    {
      "epoch": 0.3547493521579841,
      "grad_norm": 0.3542814552783966,
      "learning_rate": 2e-05,
      "loss": 0.688,
      "step": 3970
    },
    {
      "epoch": 0.3556429273523367,
      "grad_norm": 0.3793947994709015,
      "learning_rate": 2e-05,
      "loss": 0.7086,
      "step": 3980
    },
    {
      "epoch": 0.3565365025466893,
      "grad_norm": 0.4207400381565094,
      "learning_rate": 2e-05,
      "loss": 0.7057,
      "step": 3990
    },
    {
      "epoch": 0.3574300777410419,
      "grad_norm": 0.373004287481308,
      "learning_rate": 2e-05,
      "loss": 0.68,
      "step": 4000
    },
    {
      "epoch": 0.35832365293539453,
      "grad_norm": 0.4065035879611969,
      "learning_rate": 2e-05,
      "loss": 0.6917,
      "step": 4010
    },
    {
      "epoch": 0.3592172281297471,
      "grad_norm": 0.3997182846069336,
      "learning_rate": 2e-05,
      "loss": 0.6863,
      "step": 4020
    },
    {
      "epoch": 0.3601108033240997,
      "grad_norm": 0.3566669225692749,
      "learning_rate": 2e-05,
      "loss": 0.6734,
      "step": 4030
    },
    {
      "epoch": 0.36100437851845235,
      "grad_norm": 0.3704060912132263,
      "learning_rate": 2e-05,
      "loss": 0.6712,
      "step": 4040
    },
    {
      "epoch": 0.3618979537128049,
      "grad_norm": 0.35917559266090393,
      "learning_rate": 2e-05,
      "loss": 0.6796,
      "step": 4050
    },
    {
      "epoch": 0.36279152890715755,
      "grad_norm": 0.3765454888343811,
      "learning_rate": 2e-05,
      "loss": 0.6951,
      "step": 4060
    },
    {
      "epoch": 0.3636851041015101,
      "grad_norm": 0.3873068690299988,
      "learning_rate": 2e-05,
      "loss": 0.6772,
      "step": 4070
    },
    {
      "epoch": 0.36457867929586274,
      "grad_norm": 0.45908427238464355,
      "learning_rate": 2e-05,
      "loss": 0.6765,
      "step": 4080
    },
    {
      "epoch": 0.36547225449021536,
      "grad_norm": 0.3917163610458374,
      "learning_rate": 2e-05,
      "loss": 0.6787,
      "step": 4090
    },
    {
      "epoch": 0.36636582968456793,
      "grad_norm": 0.39318227767944336,
      "learning_rate": 2e-05,
      "loss": 0.6915,
      "step": 4100
    },
    {
      "epoch": 0.36725940487892056,
      "grad_norm": 0.4183279573917389,
      "learning_rate": 2e-05,
      "loss": 0.6869,
      "step": 4110
    },
    {
      "epoch": 0.3681529800732732,
      "grad_norm": 0.3996908664703369,
      "learning_rate": 2e-05,
      "loss": 0.7152,
      "step": 4120
    },
    {
      "epoch": 0.36904655526762575,
      "grad_norm": 0.38322579860687256,
      "learning_rate": 2e-05,
      "loss": 0.6713,
      "step": 4130
    },
    {
      "epoch": 0.3699401304619784,
      "grad_norm": 0.417958527803421,
      "learning_rate": 2e-05,
      "loss": 0.6647,
      "step": 4140
    },
    {
      "epoch": 0.370833705656331,
      "grad_norm": 0.386135458946228,
      "learning_rate": 2e-05,
      "loss": 0.6773,
      "step": 4150
    },
    {
      "epoch": 0.3717272808506836,
      "grad_norm": 0.3646849989891052,
      "learning_rate": 2e-05,
      "loss": 0.6784,
      "step": 4160
    },
    {
      "epoch": 0.3726208560450362,
      "grad_norm": 0.38974642753601074,
      "learning_rate": 2e-05,
      "loss": 0.6961,
      "step": 4170
    },
    {
      "epoch": 0.3735144312393888,
      "grad_norm": 0.36325088143348694,
      "learning_rate": 2e-05,
      "loss": 0.6646,
      "step": 4180
    },
    {
      "epoch": 0.3744080064337414,
      "grad_norm": 0.3774178624153137,
      "learning_rate": 2e-05,
      "loss": 0.6642,
      "step": 4190
    },
    {
      "epoch": 0.375301581628094,
      "grad_norm": 0.3733026385307312,
      "learning_rate": 2e-05,
      "loss": 0.6835,
      "step": 4200
    },
    {
      "epoch": 0.3761951568224466,
      "grad_norm": 0.4059685170650482,
      "learning_rate": 2e-05,
      "loss": 0.7016,
      "step": 4210
    },
    {
      "epoch": 0.3770887320167992,
      "grad_norm": 0.3935318887233734,
      "learning_rate": 2e-05,
      "loss": 0.7123,
      "step": 4220
    },
    {
      "epoch": 0.37798230721115184,
      "grad_norm": 0.3663133382797241,
      "learning_rate": 2e-05,
      "loss": 0.6721,
      "step": 4230
    },
    {
      "epoch": 0.3788758824055044,
      "grad_norm": 0.3684714138507843,
      "learning_rate": 2e-05,
      "loss": 0.6795,
      "step": 4240
    },
    {
      "epoch": 0.37976945759985703,
      "grad_norm": 0.37538403272628784,
      "learning_rate": 2e-05,
      "loss": 0.6752,
      "step": 4250
    },
    {
      "epoch": 0.38066303279420965,
      "grad_norm": 0.4001816511154175,
      "learning_rate": 2e-05,
      "loss": 0.7118,
      "step": 4260
    },
    {
      "epoch": 0.3815566079885622,
      "grad_norm": 0.3811742663383484,
      "learning_rate": 2e-05,
      "loss": 0.6877,
      "step": 4270
    },
    {
      "epoch": 0.38245018318291485,
      "grad_norm": 0.46673205494880676,
      "learning_rate": 2e-05,
      "loss": 0.7128,
      "step": 4280
    },
    {
      "epoch": 0.3833437583772675,
      "grad_norm": 0.37563374638557434,
      "learning_rate": 2e-05,
      "loss": 0.6742,
      "step": 4290
    },
    {
      "epoch": 0.38423733357162004,
      "grad_norm": 6.816838264465332,
      "learning_rate": 2e-05,
      "loss": 0.6968,
      "step": 4300
    },
    {
      "epoch": 0.38513090876597267,
      "grad_norm": 0.36754029989242554,
      "learning_rate": 2e-05,
      "loss": 0.6976,
      "step": 4310
    },
    {
      "epoch": 0.38602448396032524,
      "grad_norm": 0.3981974720954895,
      "learning_rate": 2e-05,
      "loss": 0.6844,
      "step": 4320
    },
    {
      "epoch": 0.38691805915467786,
      "grad_norm": 0.6864988803863525,
      "learning_rate": 2e-05,
      "loss": 0.6647,
      "step": 4330
    },
    {
      "epoch": 0.3878116343490305,
      "grad_norm": 0.38520216941833496,
      "learning_rate": 2e-05,
      "loss": 0.7092,
      "step": 4340
    },
    {
      "epoch": 0.38870520954338306,
      "grad_norm": 0.3747800886631012,
      "learning_rate": 2e-05,
      "loss": 0.6966,
      "step": 4350
    },
    {
      "epoch": 0.3895987847377357,
      "grad_norm": 0.36001715064048767,
      "learning_rate": 2e-05,
      "loss": 0.6479,
      "step": 4360
    },
    {
      "epoch": 0.3904923599320883,
      "grad_norm": 0.35367342829704285,
      "learning_rate": 2e-05,
      "loss": 0.6809,
      "step": 4370
    },
    {
      "epoch": 0.3913859351264409,
      "grad_norm": 0.39095789194107056,
      "learning_rate": 2e-05,
      "loss": 0.6878,
      "step": 4380
    },
    {
      "epoch": 0.3922795103207935,
      "grad_norm": 0.3698689043521881,
      "learning_rate": 2e-05,
      "loss": 0.6823,
      "step": 4390
    },
    {
      "epoch": 0.3931730855151461,
      "grad_norm": 0.37408924102783203,
      "learning_rate": 2e-05,
      "loss": 0.6906,
      "step": 4400
    },
    {
      "epoch": 0.3940666607094987,
      "grad_norm": 0.36290866136550903,
      "learning_rate": 2e-05,
      "loss": 0.6966,
      "step": 4410
    },
    {
      "epoch": 0.3949602359038513,
      "grad_norm": 0.364419549703598,
      "learning_rate": 2e-05,
      "loss": 0.743,
      "step": 4420
    },
    {
      "epoch": 0.3958538110982039,
      "grad_norm": 0.35897934436798096,
      "learning_rate": 2e-05,
      "loss": 0.6659,
      "step": 4430
    },
    {
      "epoch": 0.3967473862925565,
      "grad_norm": 0.34913840889930725,
      "learning_rate": 2e-05,
      "loss": 0.6977,
      "step": 4440
    },
    {
      "epoch": 0.39764096148690914,
      "grad_norm": 0.36154356598854065,
      "learning_rate": 2e-05,
      "loss": 0.7255,
      "step": 4450
    },
    {
      "epoch": 0.3985345366812617,
      "grad_norm": 0.4215991497039795,
      "learning_rate": 2e-05,
      "loss": 0.6895,
      "step": 4460
    },
    {
      "epoch": 0.39942811187561433,
      "grad_norm": 0.519006073474884,
      "learning_rate": 2e-05,
      "loss": 0.6906,
      "step": 4470
    },
    {
      "epoch": 0.40032168706996696,
      "grad_norm": 0.36556845903396606,
      "learning_rate": 2e-05,
      "loss": 0.679,
      "step": 4480
    },
    {
      "epoch": 0.4012152622643195,
      "grad_norm": 0.3858962059020996,
      "learning_rate": 2e-05,
      "loss": 0.7275,
      "step": 4490
    },
    {
      "epoch": 0.40210883745867215,
      "grad_norm": 0.3571610152721405,
      "learning_rate": 2e-05,
      "loss": 0.6791,
      "step": 4500
    },
    {
      "epoch": 0.4030024126530248,
      "grad_norm": 0.3609004318714142,
      "learning_rate": 2e-05,
      "loss": 0.6788,
      "step": 4510
    },
    {
      "epoch": 0.40389598784737735,
      "grad_norm": 0.37380915880203247,
      "learning_rate": 2e-05,
      "loss": 0.6339,
      "step": 4520
    },
    {
      "epoch": 0.40478956304172997,
      "grad_norm": 0.49198201298713684,
      "learning_rate": 2e-05,
      "loss": 0.6523,
      "step": 4530
    },
    {
      "epoch": 0.40568313823608254,
      "grad_norm": 0.3402051329612732,
      "learning_rate": 2e-05,
      "loss": 0.69,
      "step": 4540
    },
    {
      "epoch": 0.40657671343043517,
      "grad_norm": 0.40370455384254456,
      "learning_rate": 2e-05,
      "loss": 0.6949,
      "step": 4550
    },
    {
      "epoch": 0.4074702886247878,
      "grad_norm": 0.3506670594215393,
      "learning_rate": 2e-05,
      "loss": 0.6948,
      "step": 4560
    },
    {
      "epoch": 0.40836386381914036,
      "grad_norm": 0.363017201423645,
      "learning_rate": 2e-05,
      "loss": 0.6584,
      "step": 4570
    },
    {
      "epoch": 0.409257439013493,
      "grad_norm": 0.38264966011047363,
      "learning_rate": 2e-05,
      "loss": 0.7156,
      "step": 4580
    },
    {
      "epoch": 0.4101510142078456,
      "grad_norm": 0.3690856099128723,
      "learning_rate": 2e-05,
      "loss": 0.6768,
      "step": 4590
    },
    {
      "epoch": 0.4110445894021982,
      "grad_norm": 0.599280595779419,
      "learning_rate": 2e-05,
      "loss": 0.6952,
      "step": 4600
    },
    {
      "epoch": 0.4119381645965508,
      "grad_norm": 0.3705596923828125,
      "learning_rate": 2e-05,
      "loss": 0.6967,
      "step": 4610
    },
    {
      "epoch": 0.41283173979090343,
      "grad_norm": 0.36947357654571533,
      "learning_rate": 2e-05,
      "loss": 0.6918,
      "step": 4620
    },
    {
      "epoch": 0.413725314985256,
      "grad_norm": 0.38991454243659973,
      "learning_rate": 2e-05,
      "loss": 0.6834,
      "step": 4630
    },
    {
      "epoch": 0.4146188901796086,
      "grad_norm": 0.3343438506126404,
      "learning_rate": 2e-05,
      "loss": 0.6647,
      "step": 4640
    },
    {
      "epoch": 0.4155124653739612,
      "grad_norm": 0.3949652314186096,
      "learning_rate": 2e-05,
      "loss": 0.6711,
      "step": 4650
    },
    {
      "epoch": 0.4164060405683138,
      "grad_norm": 0.37002232670783997,
      "learning_rate": 2e-05,
      "loss": 0.6903,
      "step": 4660
    },
    {
      "epoch": 0.41729961576266644,
      "grad_norm": 0.383247971534729,
      "learning_rate": 2e-05,
      "loss": 0.6945,
      "step": 4670
    },
    {
      "epoch": 0.418193190957019,
      "grad_norm": 0.38935351371765137,
      "learning_rate": 2e-05,
      "loss": 0.6894,
      "step": 4680
    },
    {
      "epoch": 0.41908676615137164,
      "grad_norm": 0.38259515166282654,
      "learning_rate": 2e-05,
      "loss": 0.6986,
      "step": 4690
    },
    {
      "epoch": 0.41998034134572426,
      "grad_norm": 0.348149836063385,
      "learning_rate": 2e-05,
      "loss": 0.6872,
      "step": 4700
    },
    {
      "epoch": 0.42087391654007683,
      "grad_norm": 0.36275845766067505,
      "learning_rate": 2e-05,
      "loss": 0.6837,
      "step": 4710
    },
    {
      "epoch": 0.42176749173442946,
      "grad_norm": 0.4239656329154968,
      "learning_rate": 2e-05,
      "loss": 0.7012,
      "step": 4720
    },
    {
      "epoch": 0.4226610669287821,
      "grad_norm": 0.3658117651939392,
      "learning_rate": 2e-05,
      "loss": 0.7016,
      "step": 4730
    },
    {
      "epoch": 0.42355464212313465,
      "grad_norm": 0.377859503030777,
      "learning_rate": 2e-05,
      "loss": 0.6828,
      "step": 4740
    },
    {
      "epoch": 0.4244482173174873,
      "grad_norm": 0.38955992460250854,
      "learning_rate": 2e-05,
      "loss": 0.6773,
      "step": 4750
    },
    {
      "epoch": 0.42534179251183984,
      "grad_norm": 0.3484950065612793,
      "learning_rate": 2e-05,
      "loss": 0.7054,
      "step": 4760
    },
    {
      "epoch": 0.42623536770619247,
      "grad_norm": 0.3950771689414978,
      "learning_rate": 2e-05,
      "loss": 0.6854,
      "step": 4770
    },
    {
      "epoch": 0.4271289429005451,
      "grad_norm": 0.384011447429657,
      "learning_rate": 2e-05,
      "loss": 0.6921,
      "step": 4780
    },
    {
      "epoch": 0.42802251809489766,
      "grad_norm": 0.3784692883491516,
      "learning_rate": 2e-05,
      "loss": 0.6886,
      "step": 4790
    },
    {
      "epoch": 0.4289160932892503,
      "grad_norm": 0.3680237829685211,
      "learning_rate": 2e-05,
      "loss": 0.6835,
      "step": 4800
    },
    {
      "epoch": 0.4298096684836029,
      "grad_norm": 0.34581565856933594,
      "learning_rate": 2e-05,
      "loss": 0.6733,
      "step": 4810
    },
    {
      "epoch": 0.4307032436779555,
      "grad_norm": 0.35307255387306213,
      "learning_rate": 2e-05,
      "loss": 0.6833,
      "step": 4820
    },
    {
      "epoch": 0.4315968188723081,
      "grad_norm": 0.3662859797477722,
      "learning_rate": 2e-05,
      "loss": 0.681,
      "step": 4830
    },
    {
      "epoch": 0.43249039406666073,
      "grad_norm": 0.4486631155014038,
      "learning_rate": 2e-05,
      "loss": 0.6872,
      "step": 4840
    },
    {
      "epoch": 0.4333839692610133,
      "grad_norm": 0.3616761565208435,
      "learning_rate": 2e-05,
      "loss": 0.648,
      "step": 4850
    },
    {
      "epoch": 0.4342775444553659,
      "grad_norm": 0.3854536712169647,
      "learning_rate": 2e-05,
      "loss": 0.6601,
      "step": 4860
    },
    {
      "epoch": 0.4351711196497185,
      "grad_norm": 0.3570791780948639,
      "learning_rate": 2e-05,
      "loss": 0.7085,
      "step": 4870
    },
    {
      "epoch": 0.4360646948440711,
      "grad_norm": 0.3633802831172943,
      "learning_rate": 2e-05,
      "loss": 0.6874,
      "step": 4880
    },
    {
      "epoch": 0.43695827003842375,
      "grad_norm": 0.3554231524467468,
      "learning_rate": 2e-05,
      "loss": 0.6654,
      "step": 4890
    },
    {
      "epoch": 0.4378518452327763,
      "grad_norm": 0.5552303194999695,
      "learning_rate": 2e-05,
      "loss": 0.6903,
      "step": 4900
    },
    {
      "epoch": 0.43874542042712894,
      "grad_norm": 0.38484469056129456,
      "learning_rate": 2e-05,
      "loss": 0.6819,
      "step": 4910
    },
    {
      "epoch": 0.43963899562148157,
      "grad_norm": 0.4366731345653534,
      "learning_rate": 2e-05,
      "loss": 0.6987,
      "step": 4920
    },
    {
      "epoch": 0.44053257081583413,
      "grad_norm": 0.3423648774623871,
      "learning_rate": 2e-05,
      "loss": 0.6741,
      "step": 4930
    },
    {
      "epoch": 0.44142614601018676,
      "grad_norm": 0.3649511933326721,
      "learning_rate": 2e-05,
      "loss": 0.6907,
      "step": 4940
    },
    {
      "epoch": 0.4423197212045394,
      "grad_norm": 0.3819730281829834,
      "learning_rate": 2e-05,
      "loss": 0.6709,
      "step": 4950
    },
    {
      "epoch": 0.44321329639889195,
      "grad_norm": 0.365536630153656,
      "learning_rate": 2e-05,
      "loss": 0.7045,
      "step": 4960
    },
    {
      "epoch": 0.4441068715932446,
      "grad_norm": 0.37038612365722656,
      "learning_rate": 2e-05,
      "loss": 0.6797,
      "step": 4970
    },
    {
      "epoch": 0.4450004467875972,
      "grad_norm": 0.3828907608985901,
      "learning_rate": 2e-05,
      "loss": 0.6634,
      "step": 4980
    },
    {
      "epoch": 0.4458940219819498,
      "grad_norm": 0.38271892070770264,
      "learning_rate": 2e-05,
      "loss": 0.6289,
      "step": 4990
    },
    {
      "epoch": 0.4467875971763024,
      "grad_norm": 0.3896707594394684,
      "learning_rate": 2e-05,
      "loss": 0.7,
      "step": 5000
    },
    {
      "epoch": 0.44768117237065497,
      "grad_norm": 0.38165053725242615,
      "learning_rate": 2e-05,
      "loss": 0.656,
      "step": 5010
    },
    {
      "epoch": 0.4485747475650076,
      "grad_norm": 0.3564092218875885,
      "learning_rate": 2e-05,
      "loss": 0.6423,
      "step": 5020
    },
    {
      "epoch": 0.4494683227593602,
      "grad_norm": 0.3507450819015503,
      "learning_rate": 2e-05,
      "loss": 0.6583,
      "step": 5030
    },
    {
      "epoch": 0.4503618979537128,
      "grad_norm": 0.36991018056869507,
      "learning_rate": 2e-05,
      "loss": 0.685,
      "step": 5040
    },
    {
      "epoch": 0.4512554731480654,
      "grad_norm": 0.37723660469055176,
      "learning_rate": 2e-05,
      "loss": 0.6693,
      "step": 5050
    },
    {
      "epoch": 0.45214904834241804,
      "grad_norm": 0.3848140239715576,
      "learning_rate": 2e-05,
      "loss": 0.6759,
      "step": 5060
    },
    {
      "epoch": 0.4530426235367706,
      "grad_norm": 0.3824312388896942,
      "learning_rate": 2e-05,
      "loss": 0.6597,
      "step": 5070
    },
    {
      "epoch": 0.45393619873112323,
      "grad_norm": 0.3876231908798218,
      "learning_rate": 2e-05,
      "loss": 0.6932,
      "step": 5080
    },
    {
      "epoch": 0.45482977392547586,
      "grad_norm": 0.4114287197589874,
      "learning_rate": 2e-05,
      "loss": 0.6498,
      "step": 5090
    },
    {
      "epoch": 0.4557233491198284,
      "grad_norm": 0.3386726677417755,
      "learning_rate": 2e-05,
      "loss": 0.68,
      "step": 5100
    },
    {
      "epoch": 0.45661692431418105,
      "grad_norm": 0.40712520480155945,
      "learning_rate": 2e-05,
      "loss": 0.7036,
      "step": 5110
    },
    {
      "epoch": 0.4575104995085336,
      "grad_norm": 0.3941819667816162,
      "learning_rate": 2e-05,
      "loss": 0.6766,
      "step": 5120
    },
    {
      "epoch": 0.45840407470288624,
      "grad_norm": 0.3632022738456726,
      "learning_rate": 2e-05,
      "loss": 0.6723,
      "step": 5130
    },
    {
      "epoch": 0.45929764989723887,
      "grad_norm": 0.3453669548034668,
      "learning_rate": 2e-05,
      "loss": 0.6795,
      "step": 5140
    },
    {
      "epoch": 0.46019122509159144,
      "grad_norm": 0.3640795052051544,
      "learning_rate": 2e-05,
      "loss": 0.6685,
      "step": 5150
    },
    {
      "epoch": 0.46108480028594406,
      "grad_norm": 0.3645226061344147,
      "learning_rate": 2e-05,
      "loss": 0.6857,
      "step": 5160
    },
    {
      "epoch": 0.4619783754802967,
      "grad_norm": 0.39559850096702576,
      "learning_rate": 2e-05,
      "loss": 0.6907,
      "step": 5170
    },
    {
      "epoch": 0.46287195067464926,
      "grad_norm": 0.3610205352306366,
      "learning_rate": 2e-05,
      "loss": 0.701,
      "step": 5180
    },
    {
      "epoch": 0.4637655258690019,
      "grad_norm": 0.37898167967796326,
      "learning_rate": 2e-05,
      "loss": 0.6801,
      "step": 5190
    },
    {
      "epoch": 0.4646591010633545,
      "grad_norm": 0.7604682445526123,
      "learning_rate": 2e-05,
      "loss": 0.697,
      "step": 5200
    },
    {
      "epoch": 0.4655526762577071,
      "grad_norm": 0.38393741846084595,
      "learning_rate": 2e-05,
      "loss": 0.6868,
      "step": 5210
    },
    {
      "epoch": 0.4664462514520597,
      "grad_norm": 0.42327502369880676,
      "learning_rate": 2e-05,
      "loss": 0.6832,
      "step": 5220
    },
    {
      "epoch": 0.46733982664641227,
      "grad_norm": 0.4091464877128601,
      "learning_rate": 2e-05,
      "loss": 0.6873,
      "step": 5230
    },
    {
      "epoch": 0.4682334018407649,
      "grad_norm": 0.3564821183681488,
      "learning_rate": 2e-05,
      "loss": 0.6697,
      "step": 5240
    },
    {
      "epoch": 0.4691269770351175,
      "grad_norm": 0.42019420862197876,
      "learning_rate": 2e-05,
      "loss": 0.6652,
      "step": 5250
    },
    {
      "epoch": 0.4700205522294701,
      "grad_norm": 0.3875599205493927,
      "learning_rate": 2e-05,
      "loss": 0.6855,
      "step": 5260
    },
    {
      "epoch": 0.4709141274238227,
      "grad_norm": 0.327510267496109,
      "learning_rate": 2e-05,
      "loss": 0.6669,
      "step": 5270
    },
    {
      "epoch": 0.47180770261817534,
      "grad_norm": 0.3666091859340668,
      "learning_rate": 2e-05,
      "loss": 0.6726,
      "step": 5280
    },
    {
      "epoch": 0.4727012778125279,
      "grad_norm": 0.38704097270965576,
      "learning_rate": 2e-05,
      "loss": 0.6669,
      "step": 5290
    },
    {
      "epoch": 0.47359485300688053,
      "grad_norm": 0.3735831677913666,
      "learning_rate": 2e-05,
      "loss": 0.6605,
      "step": 5300
    },
    {
      "epoch": 0.47448842820123316,
      "grad_norm": 0.3813984990119934,
      "learning_rate": 2e-05,
      "loss": 0.666,
      "step": 5310
    },
    {
      "epoch": 0.47538200339558573,
      "grad_norm": 0.3698085844516754,
      "learning_rate": 2e-05,
      "loss": 0.6879,
      "step": 5320
    },
    {
      "epoch": 0.47627557858993835,
      "grad_norm": 0.3574400544166565,
      "learning_rate": 2e-05,
      "loss": 0.6604,
      "step": 5330
    },
    {
      "epoch": 0.4771691537842909,
      "grad_norm": 0.34233132004737854,
      "learning_rate": 2e-05,
      "loss": 0.6731,
      "step": 5340
    },
    {
      "epoch": 0.47806272897864355,
      "grad_norm": 0.34131020307540894,
      "learning_rate": 2e-05,
      "loss": 0.6784,
      "step": 5350
    },
    {
      "epoch": 0.4789563041729962,
      "grad_norm": 0.40764012932777405,
      "learning_rate": 2e-05,
      "loss": 0.6881,
      "step": 5360
    },
    {
      "epoch": 0.47984987936734874,
      "grad_norm": 0.38132914900779724,
      "learning_rate": 2e-05,
      "loss": 0.6773,
      "step": 5370
    },
    {
      "epoch": 0.48074345456170137,
      "grad_norm": 0.3372586667537689,
      "learning_rate": 2e-05,
      "loss": 0.6627,
      "step": 5380
    },
    {
      "epoch": 0.481637029756054,
      "grad_norm": 0.36021411418914795,
      "learning_rate": 2e-05,
      "loss": 0.6692,
      "step": 5390
    },
    {
      "epoch": 0.48253060495040656,
      "grad_norm": 0.34522297978401184,
      "learning_rate": 2e-05,
      "loss": 0.6643,
      "step": 5400
    },
    {
      "epoch": 0.4834241801447592,
      "grad_norm": 0.3908405005931854,
      "learning_rate": 2e-05,
      "loss": 0.7312,
      "step": 5410
    },
    {
      "epoch": 0.4843177553391118,
      "grad_norm": 0.3654153048992157,
      "learning_rate": 2e-05,
      "loss": 0.6682,
      "step": 5420
    },
    {
      "epoch": 0.4852113305334644,
      "grad_norm": 0.3864952325820923,
      "learning_rate": 2e-05,
      "loss": 0.7058,
      "step": 5430
    },
    {
      "epoch": 0.486104905727817,
      "grad_norm": 0.35901758074760437,
      "learning_rate": 2e-05,
      "loss": 0.6591,
      "step": 5440
    },
    {
      "epoch": 0.4869984809221696,
      "grad_norm": 0.36962005496025085,
      "learning_rate": 2e-05,
      "loss": 0.7032,
      "step": 5450
    },
    {
      "epoch": 0.4878920561165222,
      "grad_norm": 0.4108622968196869,
      "learning_rate": 2e-05,
      "loss": 0.6663,
      "step": 5460
    },
    {
      "epoch": 0.4887856313108748,
      "grad_norm": 0.3410824239253998,
      "learning_rate": 2e-05,
      "loss": 0.6652,
      "step": 5470
    },
    {
      "epoch": 0.4896792065052274,
      "grad_norm": 0.4414083659648895,
      "learning_rate": 2e-05,
      "loss": 0.6702,
      "step": 5480
    },
    {
      "epoch": 0.49057278169958,
      "grad_norm": 0.3652523159980774,
      "learning_rate": 2e-05,
      "loss": 0.6707,
      "step": 5490
    },
    {
      "epoch": 0.49146635689393264,
      "grad_norm": 0.3445082902908325,
      "learning_rate": 2e-05,
      "loss": 0.6959,
      "step": 5500
    },
    {
      "epoch": 0.4923599320882852,
      "grad_norm": 0.3741017282009125,
      "learning_rate": 2e-05,
      "loss": 0.6526,
      "step": 5510
    },
    {
      "epoch": 0.49325350728263784,
      "grad_norm": 0.36958327889442444,
      "learning_rate": 2e-05,
      "loss": 0.6867,
      "step": 5520
    },
    {
      "epoch": 0.49414708247699046,
      "grad_norm": 0.3438335657119751,
      "learning_rate": 2e-05,
      "loss": 0.6798,
      "step": 5530
    },
    {
      "epoch": 0.49504065767134303,
      "grad_norm": 0.3711438477039337,
      "learning_rate": 2e-05,
      "loss": 0.6731,
      "step": 5540
    },
    {
      "epoch": 0.49593423286569566,
      "grad_norm": 0.40566444396972656,
      "learning_rate": 2e-05,
      "loss": 0.6639,
      "step": 5550
    },
    {
      "epoch": 0.4968278080600482,
      "grad_norm": 0.34888753294944763,
      "learning_rate": 2e-05,
      "loss": 0.6791,
      "step": 5560
    },
    {
      "epoch": 0.49772138325440085,
      "grad_norm": 0.3506530225276947,
      "learning_rate": 2e-05,
      "loss": 0.6366,
      "step": 5570
    },
    {
      "epoch": 0.4986149584487535,
      "grad_norm": 0.3712237775325775,
      "learning_rate": 2e-05,
      "loss": 0.675,
      "step": 5580
    },
    {
      "epoch": 0.49950853364310605,
      "grad_norm": 0.3795869052410126,
      "learning_rate": 2e-05,
      "loss": 0.6804,
      "step": 5590
    },
    {
      "epoch": 0.5004021088374587,
      "grad_norm": 0.3536120355129242,
      "learning_rate": 2e-05,
      "loss": 0.6728,
      "step": 5600
    },
    {
      "epoch": 0.5012956840318112,
      "grad_norm": 0.39703911542892456,
      "learning_rate": 2e-05,
      "loss": 0.6911,
      "step": 5610
    },
    {
      "epoch": 0.5021892592261639,
      "grad_norm": 0.3442331552505493,
      "learning_rate": 2e-05,
      "loss": 0.6793,
      "step": 5620
    },
    {
      "epoch": 0.5030828344205165,
      "grad_norm": 0.35264086723327637,
      "learning_rate": 2e-05,
      "loss": 0.6757,
      "step": 5630
    },
    {
      "epoch": 0.5039764096148691,
      "grad_norm": 0.3849450349807739,
      "learning_rate": 2e-05,
      "loss": 0.7009,
      "step": 5640
    },
    {
      "epoch": 0.5048699848092217,
      "grad_norm": 0.3662022352218628,
      "learning_rate": 2e-05,
      "loss": 0.6531,
      "step": 5650
    },
    {
      "epoch": 0.5057635600035743,
      "grad_norm": 0.36823198199272156,
      "learning_rate": 2e-05,
      "loss": 0.6905,
      "step": 5660
    },
    {
      "epoch": 0.5066571351979269,
      "grad_norm": 0.36183229088783264,
      "learning_rate": 2e-05,
      "loss": 0.6885,
      "step": 5670
    },
    {
      "epoch": 0.5075507103922795,
      "grad_norm": 0.3490225374698639,
      "learning_rate": 2e-05,
      "loss": 0.6583,
      "step": 5680
    },
    {
      "epoch": 0.5084442855866321,
      "grad_norm": 0.3835480213165283,
      "learning_rate": 2e-05,
      "loss": 0.6364,
      "step": 5690
    },
    {
      "epoch": 0.5093378607809848,
      "grad_norm": 0.3251011073589325,
      "learning_rate": 2e-05,
      "loss": 0.6848,
      "step": 5700
    },
    {
      "epoch": 0.5102314359753374,
      "grad_norm": 0.33365899324417114,
      "learning_rate": 2e-05,
      "loss": 0.6681,
      "step": 5710
    },
    {
      "epoch": 0.5111250111696899,
      "grad_norm": 0.43802380561828613,
      "learning_rate": 2e-05,
      "loss": 0.7042,
      "step": 5720
    },
    {
      "epoch": 0.5120185863640425,
      "grad_norm": 0.3589516580104828,
      "learning_rate": 2e-05,
      "loss": 0.7101,
      "step": 5730
    },
    {
      "epoch": 0.5129121615583951,
      "grad_norm": 0.3624807894229889,
      "learning_rate": 2e-05,
      "loss": 0.6979,
      "step": 5740
    },
    {
      "epoch": 0.5138057367527478,
      "grad_norm": 0.4105142652988434,
      "learning_rate": 2e-05,
      "loss": 0.6885,
      "step": 5750
    },
    {
      "epoch": 0.5146993119471004,
      "grad_norm": 0.4077311158180237,
      "learning_rate": 2e-05,
      "loss": 0.6976,
      "step": 5760
    },
    {
      "epoch": 0.5155928871414529,
      "grad_norm": 0.38422033190727234,
      "learning_rate": 2e-05,
      "loss": 0.6549,
      "step": 5770
    },
    {
      "epoch": 0.5164864623358055,
      "grad_norm": 0.36182090640068054,
      "learning_rate": 2e-05,
      "loss": 0.6722,
      "step": 5780
    },
    {
      "epoch": 0.5173800375301582,
      "grad_norm": 0.35763514041900635,
      "learning_rate": 2e-05,
      "loss": 0.6756,
      "step": 5790
    },
    {
      "epoch": 0.5182736127245108,
      "grad_norm": 0.3467260003089905,
      "learning_rate": 2e-05,
      "loss": 0.6881,
      "step": 5800
    },
    {
      "epoch": 0.5191671879188634,
      "grad_norm": 0.36511150002479553,
      "learning_rate": 2e-05,
      "loss": 0.6702,
      "step": 5810
    },
    {
      "epoch": 0.520060763113216,
      "grad_norm": 0.37722599506378174,
      "learning_rate": 2e-05,
      "loss": 0.7008,
      "step": 5820
    },
    {
      "epoch": 0.5209543383075685,
      "grad_norm": 0.3883686065673828,
      "learning_rate": 2e-05,
      "loss": 0.6705,
      "step": 5830
    },
    {
      "epoch": 0.5218479135019212,
      "grad_norm": 0.41093751788139343,
      "learning_rate": 2e-05,
      "loss": 0.6694,
      "step": 5840
    },
    {
      "epoch": 0.5227414886962738,
      "grad_norm": 0.36615675687789917,
      "learning_rate": 2e-05,
      "loss": 0.6393,
      "step": 5850
    },
    {
      "epoch": 0.5236350638906264,
      "grad_norm": 0.3992595970630646,
      "learning_rate": 2e-05,
      "loss": 0.6898,
      "step": 5860
    },
    {
      "epoch": 0.524528639084979,
      "grad_norm": 0.4387722611427307,
      "learning_rate": 2e-05,
      "loss": 0.6981,
      "step": 5870
    },
    {
      "epoch": 0.5254222142793316,
      "grad_norm": 0.36492928862571716,
      "learning_rate": 2e-05,
      "loss": 0.6529,
      "step": 5880
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.3949265778064728,
      "learning_rate": 2e-05,
      "loss": 0.6747,
      "step": 5890
    },
    {
      "epoch": 0.5272093646680368,
      "grad_norm": 0.35494306683540344,
      "learning_rate": 2e-05,
      "loss": 0.6549,
      "step": 5900
    },
    {
      "epoch": 0.5281029398623894,
      "grad_norm": 0.40394827723503113,
      "learning_rate": 2e-05,
      "loss": 0.6588,
      "step": 5910
    },
    {
      "epoch": 0.5289965150567421,
      "grad_norm": 0.3773826062679291,
      "learning_rate": 2e-05,
      "loss": 0.6786,
      "step": 5920
    },
    {
      "epoch": 0.5298900902510947,
      "grad_norm": 0.38134559988975525,
      "learning_rate": 2e-05,
      "loss": 0.7167,
      "step": 5930
    },
    {
      "epoch": 0.5307836654454472,
      "grad_norm": 0.3874676823616028,
      "learning_rate": 2e-05,
      "loss": 0.6714,
      "step": 5940
    },
    {
      "epoch": 0.5316772406397998,
      "grad_norm": 0.3672949969768524,
      "learning_rate": 2e-05,
      "loss": 0.6564,
      "step": 5950
    },
    {
      "epoch": 0.5325708158341524,
      "grad_norm": 0.37121301889419556,
      "learning_rate": 2e-05,
      "loss": 0.6812,
      "step": 5960
    },
    {
      "epoch": 0.5334643910285051,
      "grad_norm": 0.3638802468776703,
      "learning_rate": 2e-05,
      "loss": 0.67,
      "step": 5970
    },
    {
      "epoch": 0.5343579662228577,
      "grad_norm": 0.37844139337539673,
      "learning_rate": 2e-05,
      "loss": 0.6973,
      "step": 5980
    },
    {
      "epoch": 0.5352515414172102,
      "grad_norm": 0.38300901651382446,
      "learning_rate": 2e-05,
      "loss": 0.6509,
      "step": 5990
    },
    {
      "epoch": 0.5361451166115628,
      "grad_norm": 0.3948492705821991,
      "learning_rate": 2e-05,
      "loss": 0.6953,
      "step": 6000
    },
    {
      "epoch": 0.5370386918059155,
      "grad_norm": 0.35126709938049316,
      "learning_rate": 2e-05,
      "loss": 0.6388,
      "step": 6010
    },
    {
      "epoch": 0.5379322670002681,
      "grad_norm": 0.34798088669776917,
      "learning_rate": 2e-05,
      "loss": 0.6564,
      "step": 6020
    },
    {
      "epoch": 0.5388258421946207,
      "grad_norm": 0.375598281621933,
      "learning_rate": 2e-05,
      "loss": 0.647,
      "step": 6030
    },
    {
      "epoch": 0.5397194173889733,
      "grad_norm": 0.3946281969547272,
      "learning_rate": 2e-05,
      "loss": 0.6576,
      "step": 6040
    },
    {
      "epoch": 0.5406129925833258,
      "grad_norm": 0.3495481014251709,
      "learning_rate": 2e-05,
      "loss": 0.6762,
      "step": 6050
    },
    {
      "epoch": 0.5415065677776785,
      "grad_norm": 0.3775424659252167,
      "learning_rate": 2e-05,
      "loss": 0.6935,
      "step": 6060
    },
    {
      "epoch": 0.5424001429720311,
      "grad_norm": 0.3790522813796997,
      "learning_rate": 2e-05,
      "loss": 0.6536,
      "step": 6070
    },
    {
      "epoch": 0.5432937181663837,
      "grad_norm": 0.37067389488220215,
      "learning_rate": 2e-05,
      "loss": 0.6752,
      "step": 6080
    },
    {
      "epoch": 0.5441872933607363,
      "grad_norm": 0.36724749207496643,
      "learning_rate": 2e-05,
      "loss": 0.6508,
      "step": 6090
    },
    {
      "epoch": 0.5450808685550889,
      "grad_norm": 0.3842104375362396,
      "learning_rate": 2e-05,
      "loss": 0.6584,
      "step": 6100
    },
    {
      "epoch": 0.5459744437494415,
      "grad_norm": 0.3364677429199219,
      "learning_rate": 2e-05,
      "loss": 0.6802,
      "step": 6110
    },
    {
      "epoch": 0.5468680189437941,
      "grad_norm": 0.36615097522735596,
      "learning_rate": 2e-05,
      "loss": 0.6511,
      "step": 6120
    },
    {
      "epoch": 0.5477615941381467,
      "grad_norm": 0.4121459126472473,
      "learning_rate": 2e-05,
      "loss": 0.7655,
      "step": 6130
    },
    {
      "epoch": 0.5486551693324994,
      "grad_norm": 0.3977929949760437,
      "learning_rate": 2e-05,
      "loss": 0.652,
      "step": 6140
    },
    {
      "epoch": 0.549548744526852,
      "grad_norm": 0.3776629865169525,
      "learning_rate": 2e-05,
      "loss": 0.6354,
      "step": 6150
    },
    {
      "epoch": 0.5504423197212045,
      "grad_norm": 0.3717542588710785,
      "learning_rate": 2e-05,
      "loss": 0.6745,
      "step": 6160
    },
    {
      "epoch": 0.5513358949155571,
      "grad_norm": 0.41286662220954895,
      "learning_rate": 2e-05,
      "loss": 0.6843,
      "step": 6170
    },
    {
      "epoch": 0.5522294701099097,
      "grad_norm": 0.40666937828063965,
      "learning_rate": 2e-05,
      "loss": 0.6457,
      "step": 6180
    },
    {
      "epoch": 0.5531230453042624,
      "grad_norm": 0.4095596373081207,
      "learning_rate": 2e-05,
      "loss": 0.6795,
      "step": 6190
    },
    {
      "epoch": 0.554016620498615,
      "grad_norm": 0.37007346749305725,
      "learning_rate": 2e-05,
      "loss": 0.6645,
      "step": 6200
    },
    {
      "epoch": 0.5549101956929675,
      "grad_norm": 0.3702617883682251,
      "learning_rate": 2e-05,
      "loss": 0.6633,
      "step": 6210
    },
    {
      "epoch": 0.5558037708873201,
      "grad_norm": 0.3570536673069,
      "learning_rate": 2e-05,
      "loss": 0.6627,
      "step": 6220
    },
    {
      "epoch": 0.5566973460816728,
      "grad_norm": 0.3929566740989685,
      "learning_rate": 2e-05,
      "loss": 0.6547,
      "step": 6230
    },
    {
      "epoch": 0.5575909212760254,
      "grad_norm": 0.36823663115501404,
      "learning_rate": 2e-05,
      "loss": 0.67,
      "step": 6240
    },
    {
      "epoch": 0.558484496470378,
      "grad_norm": 0.35621607303619385,
      "learning_rate": 2e-05,
      "loss": 0.679,
      "step": 6250
    },
    {
      "epoch": 0.5593780716647306,
      "grad_norm": 0.3739296793937683,
      "learning_rate": 2e-05,
      "loss": 0.6639,
      "step": 6260
    },
    {
      "epoch": 0.5602716468590832,
      "grad_norm": 0.3502756357192993,
      "learning_rate": 2e-05,
      "loss": 0.6444,
      "step": 6270
    },
    {
      "epoch": 0.5611652220534358,
      "grad_norm": 0.41235095262527466,
      "learning_rate": 2e-05,
      "loss": 0.6881,
      "step": 6280
    },
    {
      "epoch": 0.5620587972477884,
      "grad_norm": 0.48180878162384033,
      "learning_rate": 2e-05,
      "loss": 0.6398,
      "step": 6290
    },
    {
      "epoch": 0.562952372442141,
      "grad_norm": 0.3508879542350769,
      "learning_rate": 2e-05,
      "loss": 0.6531,
      "step": 6300
    },
    {
      "epoch": 0.5638459476364936,
      "grad_norm": 0.3862893879413605,
      "learning_rate": 2e-05,
      "loss": 0.6715,
      "step": 6310
    },
    {
      "epoch": 0.5647395228308462,
      "grad_norm": 0.36231037974357605,
      "learning_rate": 2e-05,
      "loss": 0.6466,
      "step": 6320
    },
    {
      "epoch": 0.5656330980251988,
      "grad_norm": 0.3873510956764221,
      "learning_rate": 2e-05,
      "loss": 0.663,
      "step": 6330
    },
    {
      "epoch": 0.5665266732195514,
      "grad_norm": 0.3823647201061249,
      "learning_rate": 2e-05,
      "loss": 0.6833,
      "step": 6340
    },
    {
      "epoch": 0.567420248413904,
      "grad_norm": 0.36157557368278503,
      "learning_rate": 2e-05,
      "loss": 0.6674,
      "step": 6350
    },
    {
      "epoch": 0.5683138236082567,
      "grad_norm": 0.3770220875740051,
      "learning_rate": 2e-05,
      "loss": 0.6204,
      "step": 6360
    },
    {
      "epoch": 0.5692073988026093,
      "grad_norm": 0.33984336256980896,
      "learning_rate": 2e-05,
      "loss": 0.6694,
      "step": 6370
    },
    {
      "epoch": 0.5701009739969618,
      "grad_norm": 0.3429902195930481,
      "learning_rate": 2e-05,
      "loss": 0.6656,
      "step": 6380
    },
    {
      "epoch": 0.5709945491913144,
      "grad_norm": 0.34833529591560364,
      "learning_rate": 2e-05,
      "loss": 0.6577,
      "step": 6390
    },
    {
      "epoch": 0.571888124385667,
      "grad_norm": 0.4005464017391205,
      "learning_rate": 2e-05,
      "loss": 0.6507,
      "step": 6400
    },
    {
      "epoch": 0.5727816995800197,
      "grad_norm": 0.3745936155319214,
      "learning_rate": 2e-05,
      "loss": 0.6692,
      "step": 6410
    },
    {
      "epoch": 0.5736752747743723,
      "grad_norm": 0.33218619227409363,
      "learning_rate": 2e-05,
      "loss": 0.6782,
      "step": 6420
    },
    {
      "epoch": 0.5745688499687248,
      "grad_norm": 0.36125001311302185,
      "learning_rate": 2e-05,
      "loss": 0.6316,
      "step": 6430
    },
    {
      "epoch": 0.5754624251630774,
      "grad_norm": 0.33774203062057495,
      "learning_rate": 2e-05,
      "loss": 0.6512,
      "step": 6440
    },
    {
      "epoch": 0.5763560003574301,
      "grad_norm": 0.40289124846458435,
      "learning_rate": 2e-05,
      "loss": 0.6634,
      "step": 6450
    },
    {
      "epoch": 0.5772495755517827,
      "grad_norm": 0.362293541431427,
      "learning_rate": 2e-05,
      "loss": 0.6423,
      "step": 6460
    },
    {
      "epoch": 0.5781431507461353,
      "grad_norm": 0.38059738278388977,
      "learning_rate": 2e-05,
      "loss": 0.6848,
      "step": 6470
    },
    {
      "epoch": 0.5790367259404879,
      "grad_norm": 0.37276968359947205,
      "learning_rate": 2e-05,
      "loss": 0.6822,
      "step": 6480
    },
    {
      "epoch": 0.5799303011348405,
      "grad_norm": 0.34858283400535583,
      "learning_rate": 2e-05,
      "loss": 0.6348,
      "step": 6490
    },
    {
      "epoch": 0.5808238763291931,
      "grad_norm": 0.3451462984085083,
      "learning_rate": 2e-05,
      "loss": 0.6617,
      "step": 6500
    },
    {
      "epoch": 0.5817174515235457,
      "grad_norm": 0.38942664861679077,
      "learning_rate": 2e-05,
      "loss": 0.644,
      "step": 6510
    },
    {
      "epoch": 0.5826110267178983,
      "grad_norm": 0.36882323026657104,
      "learning_rate": 2e-05,
      "loss": 0.6718,
      "step": 6520
    },
    {
      "epoch": 0.583504601912251,
      "grad_norm": 0.33940809965133667,
      "learning_rate": 2e-05,
      "loss": 0.6664,
      "step": 6530
    },
    {
      "epoch": 0.5843981771066035,
      "grad_norm": 0.3798743784427643,
      "learning_rate": 2e-05,
      "loss": 0.6491,
      "step": 6540
    },
    {
      "epoch": 0.5852917523009561,
      "grad_norm": 0.35699206590652466,
      "learning_rate": 2e-05,
      "loss": 0.6432,
      "step": 6550
    },
    {
      "epoch": 0.5861853274953087,
      "grad_norm": 0.33470699191093445,
      "learning_rate": 2e-05,
      "loss": 0.6697,
      "step": 6560
    },
    {
      "epoch": 0.5870789026896613,
      "grad_norm": 0.3633311688899994,
      "learning_rate": 2e-05,
      "loss": 0.6514,
      "step": 6570
    },
    {
      "epoch": 0.587972477884014,
      "grad_norm": 0.3314833343029022,
      "learning_rate": 2e-05,
      "loss": 0.649,
      "step": 6580
    },
    {
      "epoch": 0.5888660530783666,
      "grad_norm": 0.374553382396698,
      "learning_rate": 2e-05,
      "loss": 0.6661,
      "step": 6590
    },
    {
      "epoch": 0.5897596282727191,
      "grad_norm": 0.4215676784515381,
      "learning_rate": 2e-05,
      "loss": 0.6669,
      "step": 6600
    },
    {
      "epoch": 0.5906532034670717,
      "grad_norm": 0.329791396856308,
      "learning_rate": 2e-05,
      "loss": 0.6348,
      "step": 6610
    },
    {
      "epoch": 0.5915467786614244,
      "grad_norm": 0.41878172755241394,
      "learning_rate": 2e-05,
      "loss": 0.6704,
      "step": 6620
    },
    {
      "epoch": 0.592440353855777,
      "grad_norm": 0.36464884877204895,
      "learning_rate": 2e-05,
      "loss": 0.667,
      "step": 6630
    },
    {
      "epoch": 0.5933339290501296,
      "grad_norm": 0.3869669735431671,
      "learning_rate": 2e-05,
      "loss": 0.6477,
      "step": 6640
    },
    {
      "epoch": 0.5942275042444821,
      "grad_norm": 0.3334164023399353,
      "learning_rate": 2e-05,
      "loss": 0.6687,
      "step": 6650
    },
    {
      "epoch": 0.5951210794388347,
      "grad_norm": 0.37308958172798157,
      "learning_rate": 2e-05,
      "loss": 0.6766,
      "step": 6660
    },
    {
      "epoch": 0.5960146546331874,
      "grad_norm": 0.3769567906856537,
      "learning_rate": 2e-05,
      "loss": 0.6845,
      "step": 6670
    },
    {
      "epoch": 0.59690822982754,
      "grad_norm": 0.34867948293685913,
      "learning_rate": 2e-05,
      "loss": 0.6622,
      "step": 6680
    },
    {
      "epoch": 0.5978018050218926,
      "grad_norm": 0.37086349725723267,
      "learning_rate": 2e-05,
      "loss": 0.6958,
      "step": 6690
    },
    {
      "epoch": 0.5986953802162452,
      "grad_norm": 0.3929077088832855,
      "learning_rate": 2e-05,
      "loss": 0.671,
      "step": 6700
    },
    {
      "epoch": 0.5995889554105978,
      "grad_norm": 0.35202014446258545,
      "learning_rate": 2e-05,
      "loss": 0.6504,
      "step": 6710
    },
    {
      "epoch": 0.6004825306049504,
      "grad_norm": 0.33810484409332275,
      "learning_rate": 2e-05,
      "loss": 0.6517,
      "step": 6720
    },
    {
      "epoch": 0.601376105799303,
      "grad_norm": 0.3350893259048462,
      "learning_rate": 2e-05,
      "loss": 0.642,
      "step": 6730
    },
    {
      "epoch": 0.6022696809936556,
      "grad_norm": 0.3963231146335602,
      "learning_rate": 2e-05,
      "loss": 0.6305,
      "step": 6740
    },
    {
      "epoch": 0.6031632561880083,
      "grad_norm": 0.4189794361591339,
      "learning_rate": 2e-05,
      "loss": 0.6564,
      "step": 6750
    },
    {
      "epoch": 0.6040568313823608,
      "grad_norm": 0.38653162121772766,
      "learning_rate": 2e-05,
      "loss": 0.6941,
      "step": 6760
    },
    {
      "epoch": 0.6049504065767134,
      "grad_norm": 0.3797129988670349,
      "learning_rate": 2e-05,
      "loss": 0.6503,
      "step": 6770
    },
    {
      "epoch": 0.605843981771066,
      "grad_norm": 0.3951667249202728,
      "learning_rate": 2e-05,
      "loss": 0.6679,
      "step": 6780
    },
    {
      "epoch": 0.6067375569654186,
      "grad_norm": 0.3639308214187622,
      "learning_rate": 2e-05,
      "loss": 0.6532,
      "step": 6790
    },
    {
      "epoch": 0.6076311321597713,
      "grad_norm": 0.34797847270965576,
      "learning_rate": 2e-05,
      "loss": 0.6437,
      "step": 6800
    },
    {
      "epoch": 0.6085247073541239,
      "grad_norm": 0.3714454770088196,
      "learning_rate": 2e-05,
      "loss": 0.6559,
      "step": 6810
    },
    {
      "epoch": 0.6094182825484764,
      "grad_norm": 0.3578590154647827,
      "learning_rate": 2e-05,
      "loss": 0.682,
      "step": 6820
    },
    {
      "epoch": 0.610311857742829,
      "grad_norm": 0.3829934298992157,
      "learning_rate": 2e-05,
      "loss": 0.6485,
      "step": 6830
    },
    {
      "epoch": 0.6112054329371817,
      "grad_norm": 0.3965420126914978,
      "learning_rate": 2e-05,
      "loss": 0.6801,
      "step": 6840
    },
    {
      "epoch": 0.6120990081315343,
      "grad_norm": 0.38034310936927795,
      "learning_rate": 2e-05,
      "loss": 0.629,
      "step": 6850
    },
    {
      "epoch": 0.6129925833258869,
      "grad_norm": 0.36691030859947205,
      "learning_rate": 2e-05,
      "loss": 0.6466,
      "step": 6860
    },
    {
      "epoch": 0.6138861585202395,
      "grad_norm": 0.3570184111595154,
      "learning_rate": 2e-05,
      "loss": 0.6449,
      "step": 6870
    },
    {
      "epoch": 0.614779733714592,
      "grad_norm": 0.343162477016449,
      "learning_rate": 2e-05,
      "loss": 0.6635,
      "step": 6880
    },
    {
      "epoch": 0.6156733089089447,
      "grad_norm": 0.3828890025615692,
      "learning_rate": 2e-05,
      "loss": 0.662,
      "step": 6890
    },
    {
      "epoch": 0.6165668841032973,
      "grad_norm": 0.4234565198421478,
      "learning_rate": 2e-05,
      "loss": 0.6689,
      "step": 6900
    },
    {
      "epoch": 0.6174604592976499,
      "grad_norm": 0.35611692070961,
      "learning_rate": 2e-05,
      "loss": 0.6432,
      "step": 6910
    },
    {
      "epoch": 0.6183540344920025,
      "grad_norm": 0.3467334508895874,
      "learning_rate": 2e-05,
      "loss": 0.6532,
      "step": 6920
    },
    {
      "epoch": 0.6192476096863551,
      "grad_norm": 0.39782559871673584,
      "learning_rate": 2e-05,
      "loss": 0.654,
      "step": 6930
    },
    {
      "epoch": 0.6201411848807077,
      "grad_norm": 0.36997079849243164,
      "learning_rate": 2e-05,
      "loss": 0.6443,
      "step": 6940
    },
    {
      "epoch": 0.6210347600750603,
      "grad_norm": 0.35578516125679016,
      "learning_rate": 2e-05,
      "loss": 0.6688,
      "step": 6950
    },
    {
      "epoch": 0.6219283352694129,
      "grad_norm": 0.3609931766986847,
      "learning_rate": 2e-05,
      "loss": 0.666,
      "step": 6960
    },
    {
      "epoch": 0.6228219104637656,
      "grad_norm": 0.3672407865524292,
      "learning_rate": 2e-05,
      "loss": 0.6717,
      "step": 6970
    },
    {
      "epoch": 0.6237154856581182,
      "grad_norm": 0.6777769327163696,
      "learning_rate": 2e-05,
      "loss": 0.6535,
      "step": 6980
    },
    {
      "epoch": 0.6246090608524707,
      "grad_norm": 0.3729098439216614,
      "learning_rate": 2e-05,
      "loss": 0.6756,
      "step": 6990
    },
    {
      "epoch": 0.6255026360468233,
      "grad_norm": 0.37241581082344055,
      "learning_rate": 2e-05,
      "loss": 0.6892,
      "step": 7000
    },
    {
      "epoch": 0.626396211241176,
      "grad_norm": 0.3663364350795746,
      "learning_rate": 2e-05,
      "loss": 0.6914,
      "step": 7010
    },
    {
      "epoch": 0.6272897864355286,
      "grad_norm": 0.4170268774032593,
      "learning_rate": 2e-05,
      "loss": 0.661,
      "step": 7020
    },
    {
      "epoch": 0.6281833616298812,
      "grad_norm": 0.4091792702674866,
      "learning_rate": 2e-05,
      "loss": 0.6449,
      "step": 7030
    },
    {
      "epoch": 0.6290769368242337,
      "grad_norm": 0.36993762850761414,
      "learning_rate": 2e-05,
      "loss": 0.6832,
      "step": 7040
    },
    {
      "epoch": 0.6299705120185863,
      "grad_norm": 0.36667630076408386,
      "learning_rate": 2e-05,
      "loss": 0.65,
      "step": 7050
    },
    {
      "epoch": 0.630864087212939,
      "grad_norm": 0.37184610962867737,
      "learning_rate": 2e-05,
      "loss": 0.6655,
      "step": 7060
    },
    {
      "epoch": 0.6317576624072916,
      "grad_norm": 0.35682371258735657,
      "learning_rate": 2e-05,
      "loss": 0.648,
      "step": 7070
    },
    {
      "epoch": 0.6326512376016442,
      "grad_norm": 0.38516104221343994,
      "learning_rate": 2e-05,
      "loss": 0.6631,
      "step": 7080
    },
    {
      "epoch": 0.6335448127959968,
      "grad_norm": 0.3716648817062378,
      "learning_rate": 2e-05,
      "loss": 0.6566,
      "step": 7090
    },
    {
      "epoch": 0.6344383879903494,
      "grad_norm": 0.388762891292572,
      "learning_rate": 2e-05,
      "loss": 0.6704,
      "step": 7100
    },
    {
      "epoch": 0.635331963184702,
      "grad_norm": 0.42120128870010376,
      "learning_rate": 2e-05,
      "loss": 0.6598,
      "step": 7110
    },
    {
      "epoch": 0.6362255383790546,
      "grad_norm": 0.36003223061561584,
      "learning_rate": 2e-05,
      "loss": 0.6468,
      "step": 7120
    },
    {
      "epoch": 0.6371191135734072,
      "grad_norm": 0.32214248180389404,
      "learning_rate": 2e-05,
      "loss": 0.6368,
      "step": 7130
    },
    {
      "epoch": 0.6380126887677599,
      "grad_norm": 0.39514845609664917,
      "learning_rate": 2e-05,
      "loss": 0.6945,
      "step": 7140
    },
    {
      "epoch": 0.6389062639621124,
      "grad_norm": 0.3623945713043213,
      "learning_rate": 2e-05,
      "loss": 0.7067,
      "step": 7150
    },
    {
      "epoch": 0.639799839156465,
      "grad_norm": 0.3733201026916504,
      "learning_rate": 2e-05,
      "loss": 0.6673,
      "step": 7160
    },
    {
      "epoch": 0.6406934143508176,
      "grad_norm": 0.5380637645721436,
      "learning_rate": 2e-05,
      "loss": 0.666,
      "step": 7170
    },
    {
      "epoch": 0.6415869895451702,
      "grad_norm": 0.4173584282398224,
      "learning_rate": 2e-05,
      "loss": 0.6877,
      "step": 7180
    },
    {
      "epoch": 0.6424805647395229,
      "grad_norm": 0.37183988094329834,
      "learning_rate": 2e-05,
      "loss": 0.6252,
      "step": 7190
    },
    {
      "epoch": 0.6433741399338755,
      "grad_norm": 0.335400253534317,
      "learning_rate": 2e-05,
      "loss": 0.6866,
      "step": 7200
    },
    {
      "epoch": 0.644267715128228,
      "grad_norm": 0.4097932279109955,
      "learning_rate": 2e-05,
      "loss": 0.6758,
      "step": 7210
    },
    {
      "epoch": 0.6451612903225806,
      "grad_norm": 0.39124783873558044,
      "learning_rate": 2e-05,
      "loss": 0.6443,
      "step": 7220
    },
    {
      "epoch": 0.6460548655169333,
      "grad_norm": 0.4057658612728119,
      "learning_rate": 2e-05,
      "loss": 0.6374,
      "step": 7230
    },
    {
      "epoch": 0.6469484407112859,
      "grad_norm": 0.36567962169647217,
      "learning_rate": 2e-05,
      "loss": 0.7177,
      "step": 7240
    },
    {
      "epoch": 0.6478420159056385,
      "grad_norm": 0.369219034910202,
      "learning_rate": 2e-05,
      "loss": 0.6608,
      "step": 7250
    },
    {
      "epoch": 0.648735591099991,
      "grad_norm": 0.3781801462173462,
      "learning_rate": 2e-05,
      "loss": 0.6421,
      "step": 7260
    },
    {
      "epoch": 0.6496291662943436,
      "grad_norm": 0.38415417075157166,
      "learning_rate": 2e-05,
      "loss": 0.6485,
      "step": 7270
    },
    {
      "epoch": 0.6505227414886963,
      "grad_norm": 0.38418346643447876,
      "learning_rate": 2e-05,
      "loss": 0.6666,
      "step": 7280
    },
    {
      "epoch": 0.6514163166830489,
      "grad_norm": 0.3627326190471649,
      "learning_rate": 2e-05,
      "loss": 0.6404,
      "step": 7290
    },
    {
      "epoch": 0.6523098918774015,
      "grad_norm": 0.35553520917892456,
      "learning_rate": 2e-05,
      "loss": 0.6588,
      "step": 7300
    },
    {
      "epoch": 0.6532034670717541,
      "grad_norm": 0.372714102268219,
      "learning_rate": 2e-05,
      "loss": 0.6446,
      "step": 7310
    },
    {
      "epoch": 0.6540970422661067,
      "grad_norm": 0.3714959919452667,
      "learning_rate": 2e-05,
      "loss": 0.687,
      "step": 7320
    },
    {
      "epoch": 0.6549906174604593,
      "grad_norm": 0.37064242362976074,
      "learning_rate": 2e-05,
      "loss": 0.6663,
      "step": 7330
    },
    {
      "epoch": 0.6558841926548119,
      "grad_norm": 0.39068302512168884,
      "learning_rate": 2e-05,
      "loss": 0.6482,
      "step": 7340
    },
    {
      "epoch": 0.6567777678491645,
      "grad_norm": 0.3149115741252899,
      "learning_rate": 2e-05,
      "loss": 0.646,
      "step": 7350
    },
    {
      "epoch": 0.6576713430435172,
      "grad_norm": 0.3749960958957672,
      "learning_rate": 2e-05,
      "loss": 0.6636,
      "step": 7360
    },
    {
      "epoch": 0.6585649182378697,
      "grad_norm": 0.3658101558685303,
      "learning_rate": 2e-05,
      "loss": 0.6551,
      "step": 7370
    },
    {
      "epoch": 0.6594584934322223,
      "grad_norm": 0.34154796600341797,
      "learning_rate": 2e-05,
      "loss": 0.6865,
      "step": 7380
    },
    {
      "epoch": 0.6603520686265749,
      "grad_norm": 0.39544209837913513,
      "learning_rate": 2e-05,
      "loss": 0.65,
      "step": 7390
    },
    {
      "epoch": 0.6612456438209275,
      "grad_norm": 0.3510544002056122,
      "learning_rate": 2e-05,
      "loss": 0.6637,
      "step": 7400
    },
    {
      "epoch": 0.6621392190152802,
      "grad_norm": 0.4175526797771454,
      "learning_rate": 2e-05,
      "loss": 0.671,
      "step": 7410
    },
    {
      "epoch": 0.6630327942096328,
      "grad_norm": 0.36087992787361145,
      "learning_rate": 2e-05,
      "loss": 0.6392,
      "step": 7420
    },
    {
      "epoch": 0.6639263694039853,
      "grad_norm": 0.38082143664360046,
      "learning_rate": 2e-05,
      "loss": 0.6491,
      "step": 7430
    },
    {
      "epoch": 0.6648199445983379,
      "grad_norm": 0.36237218976020813,
      "learning_rate": 2e-05,
      "loss": 0.6729,
      "step": 7440
    },
    {
      "epoch": 0.6657135197926906,
      "grad_norm": 0.4454997181892395,
      "learning_rate": 2e-05,
      "loss": 0.6732,
      "step": 7450
    },
    {
      "epoch": 0.6666070949870432,
      "grad_norm": 0.38110703229904175,
      "learning_rate": 2e-05,
      "loss": 0.6525,
      "step": 7460
    },
    {
      "epoch": 0.6675006701813958,
      "grad_norm": 0.3655093312263489,
      "learning_rate": 2e-05,
      "loss": 0.6672,
      "step": 7470
    },
    {
      "epoch": 0.6683942453757483,
      "grad_norm": 0.3874027729034424,
      "learning_rate": 2e-05,
      "loss": 0.6777,
      "step": 7480
    },
    {
      "epoch": 0.6692878205701009,
      "grad_norm": 0.4023793637752533,
      "learning_rate": 2e-05,
      "loss": 0.6764,
      "step": 7490
    },
    {
      "epoch": 0.6701813957644536,
      "grad_norm": 0.39186614751815796,
      "learning_rate": 2e-05,
      "loss": 0.6658,
      "step": 7500
    },
    {
      "epoch": 0.6710749709588062,
      "grad_norm": 0.3666151463985443,
      "learning_rate": 2e-05,
      "loss": 0.6156,
      "step": 7510
    },
    {
      "epoch": 0.6719685461531588,
      "grad_norm": 0.38325273990631104,
      "learning_rate": 2e-05,
      "loss": 0.6584,
      "step": 7520
    },
    {
      "epoch": 0.6728621213475114,
      "grad_norm": 0.358233243227005,
      "learning_rate": 2e-05,
      "loss": 0.635,
      "step": 7530
    },
    {
      "epoch": 0.673755696541864,
      "grad_norm": 0.3498786389827728,
      "learning_rate": 2e-05,
      "loss": 0.6374,
      "step": 7540
    },
    {
      "epoch": 0.6746492717362166,
      "grad_norm": 0.3932044506072998,
      "learning_rate": 2e-05,
      "loss": 0.6846,
      "step": 7550
    },
    {
      "epoch": 0.6755428469305692,
      "grad_norm": 0.3491605818271637,
      "learning_rate": 2e-05,
      "loss": 0.6896,
      "step": 7560
    },
    {
      "epoch": 0.6764364221249218,
      "grad_norm": 0.3897663652896881,
      "learning_rate": 2e-05,
      "loss": 0.6579,
      "step": 7570
    },
    {
      "epoch": 0.6773299973192745,
      "grad_norm": 0.367182195186615,
      "learning_rate": 2e-05,
      "loss": 0.6714,
      "step": 7580
    },
    {
      "epoch": 0.678223572513627,
      "grad_norm": 0.40564435720443726,
      "learning_rate": 2e-05,
      "loss": 0.664,
      "step": 7590
    },
    {
      "epoch": 0.6791171477079796,
      "grad_norm": 0.4240153431892395,
      "learning_rate": 2e-05,
      "loss": 0.6636,
      "step": 7600
    },
    {
      "epoch": 0.6800107229023322,
      "grad_norm": 0.34123241901397705,
      "learning_rate": 2e-05,
      "loss": 0.6827,
      "step": 7610
    },
    {
      "epoch": 0.6809042980966848,
      "grad_norm": 0.3377649784088135,
      "learning_rate": 2e-05,
      "loss": 0.6278,
      "step": 7620
    },
    {
      "epoch": 0.6817978732910375,
      "grad_norm": 0.4379824101924896,
      "learning_rate": 2e-05,
      "loss": 0.6706,
      "step": 7630
    },
    {
      "epoch": 0.6826914484853901,
      "grad_norm": 0.34513503313064575,
      "learning_rate": 2e-05,
      "loss": 0.6483,
      "step": 7640
    },
    {
      "epoch": 0.6835850236797426,
      "grad_norm": 0.3698311150074005,
      "learning_rate": 2e-05,
      "loss": 0.6649,
      "step": 7650
    },
    {
      "epoch": 0.6844785988740952,
      "grad_norm": 0.38798657059669495,
      "learning_rate": 2e-05,
      "loss": 0.6477,
      "step": 7660
    },
    {
      "epoch": 0.6853721740684479,
      "grad_norm": 0.3917804956436157,
      "learning_rate": 2e-05,
      "loss": 0.6539,
      "step": 7670
    },
    {
      "epoch": 0.6862657492628005,
      "grad_norm": 0.3807283043861389,
      "learning_rate": 2e-05,
      "loss": 0.6715,
      "step": 7680
    },
    {
      "epoch": 0.6871593244571531,
      "grad_norm": 0.3713240325450897,
      "learning_rate": 2e-05,
      "loss": 0.6435,
      "step": 7690
    },
    {
      "epoch": 0.6880528996515056,
      "grad_norm": 0.3924867808818817,
      "learning_rate": 2e-05,
      "loss": 0.6543,
      "step": 7700
    },
    {
      "epoch": 0.6889464748458582,
      "grad_norm": 0.39236578345298767,
      "learning_rate": 2e-05,
      "loss": 0.6486,
      "step": 7710
    },
    {
      "epoch": 0.6898400500402109,
      "grad_norm": 0.415453165769577,
      "learning_rate": 2e-05,
      "loss": 0.6588,
      "step": 7720
    },
    {
      "epoch": 0.6907336252345635,
      "grad_norm": 0.39907023310661316,
      "learning_rate": 2e-05,
      "loss": 0.6557,
      "step": 7730
    },
    {
      "epoch": 0.6916272004289161,
      "grad_norm": 0.3817707598209381,
      "learning_rate": 2e-05,
      "loss": 0.6642,
      "step": 7740
    },
    {
      "epoch": 0.6925207756232687,
      "grad_norm": 0.3986096978187561,
      "learning_rate": 2e-05,
      "loss": 0.6447,
      "step": 7750
    },
    {
      "epoch": 0.6934143508176213,
      "grad_norm": 0.34824466705322266,
      "learning_rate": 2e-05,
      "loss": 0.6704,
      "step": 7760
    },
    {
      "epoch": 0.6943079260119739,
      "grad_norm": 0.400140643119812,
      "learning_rate": 2e-05,
      "loss": 0.642,
      "step": 7770
    },
    {
      "epoch": 0.6952015012063265,
      "grad_norm": 0.37155938148498535,
      "learning_rate": 2e-05,
      "loss": 0.623,
      "step": 7780
    },
    {
      "epoch": 0.6960950764006791,
      "grad_norm": 0.36081039905548096,
      "learning_rate": 2e-05,
      "loss": 0.6303,
      "step": 7790
    },
    {
      "epoch": 0.6969886515950318,
      "grad_norm": 0.36821597814559937,
      "learning_rate": 2e-05,
      "loss": 0.6727,
      "step": 7800
    },
    {
      "epoch": 0.6978822267893843,
      "grad_norm": 0.3827245533466339,
      "learning_rate": 2e-05,
      "loss": 0.6248,
      "step": 7810
    },
    {
      "epoch": 0.6987758019837369,
      "grad_norm": 0.3435687720775604,
      "learning_rate": 2e-05,
      "loss": 0.6455,
      "step": 7820
    },
    {
      "epoch": 0.6996693771780895,
      "grad_norm": 0.39199456572532654,
      "learning_rate": 2e-05,
      "loss": 0.6455,
      "step": 7830
    },
    {
      "epoch": 0.7005629523724421,
      "grad_norm": 0.3926967680454254,
      "learning_rate": 2e-05,
      "loss": 0.6563,
      "step": 7840
    },
    {
      "epoch": 0.7014565275667948,
      "grad_norm": 0.42693638801574707,
      "learning_rate": 2e-05,
      "loss": 0.6297,
      "step": 7850
    },
    {
      "epoch": 0.7023501027611474,
      "grad_norm": 0.36308470368385315,
      "learning_rate": 2e-05,
      "loss": 0.6355,
      "step": 7860
    },
    {
      "epoch": 0.7032436779554999,
      "grad_norm": 0.3878028988838196,
      "learning_rate": 2e-05,
      "loss": 0.6516,
      "step": 7870
    },
    {
      "epoch": 0.7041372531498525,
      "grad_norm": 0.3630359470844269,
      "learning_rate": 2e-05,
      "loss": 0.651,
      "step": 7880
    },
    {
      "epoch": 0.7050308283442052,
      "grad_norm": 0.3487136960029602,
      "learning_rate": 2e-05,
      "loss": 0.6113,
      "step": 7890
    },
    {
      "epoch": 0.7059244035385578,
      "grad_norm": 0.36301544308662415,
      "learning_rate": 2e-05,
      "loss": 0.6763,
      "step": 7900
    },
    {
      "epoch": 0.7068179787329104,
      "grad_norm": 0.3739418089389801,
      "learning_rate": 2e-05,
      "loss": 0.6592,
      "step": 7910
    },
    {
      "epoch": 0.7077115539272629,
      "grad_norm": 0.3572728931903839,
      "learning_rate": 2e-05,
      "loss": 0.6359,
      "step": 7920
    },
    {
      "epoch": 0.7086051291216156,
      "grad_norm": 0.36234769225120544,
      "learning_rate": 2e-05,
      "loss": 0.6496,
      "step": 7930
    },
    {
      "epoch": 0.7094987043159682,
      "grad_norm": 0.37305065989494324,
      "learning_rate": 2e-05,
      "loss": 0.6334,
      "step": 7940
    },
    {
      "epoch": 0.7103922795103208,
      "grad_norm": 0.33188748359680176,
      "learning_rate": 2e-05,
      "loss": 0.6222,
      "step": 7950
    },
    {
      "epoch": 0.7112858547046734,
      "grad_norm": 0.6339422464370728,
      "learning_rate": 2e-05,
      "loss": 0.654,
      "step": 7960
    },
    {
      "epoch": 0.712179429899026,
      "grad_norm": 0.38139066100120544,
      "learning_rate": 2e-05,
      "loss": 0.6345,
      "step": 7970
    },
    {
      "epoch": 0.7130730050933786,
      "grad_norm": 0.3629448115825653,
      "learning_rate": 2e-05,
      "loss": 0.6673,
      "step": 7980
    },
    {
      "epoch": 0.7139665802877312,
      "grad_norm": 0.3986387252807617,
      "learning_rate": 2e-05,
      "loss": 0.6424,
      "step": 7990
    },
    {
      "epoch": 0.7148601554820838,
      "grad_norm": 0.406455397605896,
      "learning_rate": 2e-05,
      "loss": 0.6388,
      "step": 8000
    },
    {
      "epoch": 0.7157537306764364,
      "grad_norm": 0.3624163568019867,
      "learning_rate": 2e-05,
      "loss": 0.6524,
      "step": 8010
    },
    {
      "epoch": 0.7166473058707891,
      "grad_norm": 0.3704451024532318,
      "learning_rate": 2e-05,
      "loss": 0.6546,
      "step": 8020
    },
    {
      "epoch": 0.7175408810651416,
      "grad_norm": 0.3857969045639038,
      "learning_rate": 2e-05,
      "loss": 0.6581,
      "step": 8030
    },
    {
      "epoch": 0.7184344562594942,
      "grad_norm": 0.3478555679321289,
      "learning_rate": 2e-05,
      "loss": 0.6695,
      "step": 8040
    },
    {
      "epoch": 0.7193280314538468,
      "grad_norm": 0.3627151548862457,
      "learning_rate": 2e-05,
      "loss": 0.6517,
      "step": 8050
    },
    {
      "epoch": 0.7202216066481995,
      "grad_norm": 2.139096736907959,
      "learning_rate": 2e-05,
      "loss": 0.6554,
      "step": 8060
    },
    {
      "epoch": 0.7211151818425521,
      "grad_norm": 0.3477785885334015,
      "learning_rate": 2e-05,
      "loss": 0.6367,
      "step": 8070
    },
    {
      "epoch": 0.7220087570369047,
      "grad_norm": 0.41543442010879517,
      "learning_rate": 2e-05,
      "loss": 0.6534,
      "step": 8080
    },
    {
      "epoch": 0.7229023322312572,
      "grad_norm": 0.3356950879096985,
      "learning_rate": 2e-05,
      "loss": 0.6639,
      "step": 8090
    },
    {
      "epoch": 0.7237959074256098,
      "grad_norm": 0.3489893972873688,
      "learning_rate": 2e-05,
      "loss": 0.6035,
      "step": 8100
    },
    {
      "epoch": 0.7246894826199625,
      "grad_norm": 0.3662850856781006,
      "learning_rate": 2e-05,
      "loss": 0.6617,
      "step": 8110
    },
    {
      "epoch": 0.7255830578143151,
      "grad_norm": 0.3719210624694824,
      "learning_rate": 2e-05,
      "loss": 0.6204,
      "step": 8120
    },
    {
      "epoch": 0.7264766330086677,
      "grad_norm": 0.3795759677886963,
      "learning_rate": 2e-05,
      "loss": 0.6193,
      "step": 8130
    },
    {
      "epoch": 0.7273702082030202,
      "grad_norm": 0.3709211051464081,
      "learning_rate": 2e-05,
      "loss": 0.67,
      "step": 8140
    },
    {
      "epoch": 0.7282637833973729,
      "grad_norm": 0.34277111291885376,
      "learning_rate": 2e-05,
      "loss": 0.6229,
      "step": 8150
    },
    {
      "epoch": 0.7291573585917255,
      "grad_norm": 0.3718827962875366,
      "learning_rate": 2e-05,
      "loss": 0.667,
      "step": 8160
    },
    {
      "epoch": 0.7300509337860781,
      "grad_norm": 0.4188845157623291,
      "learning_rate": 2e-05,
      "loss": 0.6383,
      "step": 8170
    },
    {
      "epoch": 0.7309445089804307,
      "grad_norm": 0.37719854712486267,
      "learning_rate": 2e-05,
      "loss": 0.6371,
      "step": 8180
    },
    {
      "epoch": 0.7318380841747834,
      "grad_norm": 0.40925782918930054,
      "learning_rate": 2e-05,
      "loss": 0.6599,
      "step": 8190
    },
    {
      "epoch": 0.7327316593691359,
      "grad_norm": 0.3855210542678833,
      "learning_rate": 2e-05,
      "loss": 0.6457,
      "step": 8200
    },
    {
      "epoch": 0.7336252345634885,
      "grad_norm": 0.4161730408668518,
      "learning_rate": 2e-05,
      "loss": 0.6208,
      "step": 8210
    },
    {
      "epoch": 0.7345188097578411,
      "grad_norm": 0.3963601589202881,
      "learning_rate": 2e-05,
      "loss": 0.65,
      "step": 8220
    },
    {
      "epoch": 0.7354123849521937,
      "grad_norm": 0.3553980588912964,
      "learning_rate": 2e-05,
      "loss": 0.6377,
      "step": 8230
    },
    {
      "epoch": 0.7363059601465464,
      "grad_norm": 0.3798747658729553,
      "learning_rate": 2e-05,
      "loss": 0.6517,
      "step": 8240
    },
    {
      "epoch": 0.7371995353408989,
      "grad_norm": 0.4044502079486847,
      "learning_rate": 2e-05,
      "loss": 0.6452,
      "step": 8250
    },
    {
      "epoch": 0.7380931105352515,
      "grad_norm": 0.3857227563858032,
      "learning_rate": 2e-05,
      "loss": 0.626,
      "step": 8260
    },
    {
      "epoch": 0.7389866857296041,
      "grad_norm": 0.3598621189594269,
      "learning_rate": 2e-05,
      "loss": 0.6135,
      "step": 8270
    },
    {
      "epoch": 0.7398802609239568,
      "grad_norm": 0.371094286441803,
      "learning_rate": 2e-05,
      "loss": 0.6468,
      "step": 8280
    },
    {
      "epoch": 0.7407738361183094,
      "grad_norm": 0.36436372995376587,
      "learning_rate": 2e-05,
      "loss": 0.6194,
      "step": 8290
    },
    {
      "epoch": 0.741667411312662,
      "grad_norm": 0.3619306683540344,
      "learning_rate": 2e-05,
      "loss": 0.6663,
      "step": 8300
    },
    {
      "epoch": 0.7425609865070145,
      "grad_norm": 0.37165531516075134,
      "learning_rate": 2e-05,
      "loss": 0.6422,
      "step": 8310
    },
    {
      "epoch": 0.7434545617013671,
      "grad_norm": 0.4265398383140564,
      "learning_rate": 2e-05,
      "loss": 0.6682,
      "step": 8320
    },
    {
      "epoch": 0.7443481368957198,
      "grad_norm": 0.3850476145744324,
      "learning_rate": 2e-05,
      "loss": 0.6445,
      "step": 8330
    },
    {
      "epoch": 0.7452417120900724,
      "grad_norm": 0.39498794078826904,
      "learning_rate": 2e-05,
      "loss": 0.6253,
      "step": 8340
    },
    {
      "epoch": 0.746135287284425,
      "grad_norm": 0.3987174332141876,
      "learning_rate": 2e-05,
      "loss": 0.6613,
      "step": 8350
    },
    {
      "epoch": 0.7470288624787776,
      "grad_norm": 0.39723101258277893,
      "learning_rate": 2e-05,
      "loss": 0.6447,
      "step": 8360
    },
    {
      "epoch": 0.7479224376731302,
      "grad_norm": 0.3784955143928528,
      "learning_rate": 2e-05,
      "loss": 0.6263,
      "step": 8370
    },
    {
      "epoch": 0.7488160128674828,
      "grad_norm": 0.3798249661922455,
      "learning_rate": 2e-05,
      "loss": 0.6126,
      "step": 8380
    },
    {
      "epoch": 0.7497095880618354,
      "grad_norm": 0.35970133543014526,
      "learning_rate": 2e-05,
      "loss": 0.6352,
      "step": 8390
    },
    {
      "epoch": 0.750603163256188,
      "grad_norm": 0.3973653316497803,
      "learning_rate": 2e-05,
      "loss": 0.6704,
      "step": 8400
    },
    {
      "epoch": 0.7514967384505407,
      "grad_norm": 0.33415666222572327,
      "learning_rate": 2e-05,
      "loss": 0.621,
      "step": 8410
    },
    {
      "epoch": 0.7523903136448932,
      "grad_norm": 0.35294389724731445,
      "learning_rate": 2e-05,
      "loss": 0.6377,
      "step": 8420
    },
    {
      "epoch": 0.7532838888392458,
      "grad_norm": 0.34990376234054565,
      "learning_rate": 2e-05,
      "loss": 0.6256,
      "step": 8430
    },
    {
      "epoch": 0.7541774640335984,
      "grad_norm": 0.3735615611076355,
      "learning_rate": 2e-05,
      "loss": 0.6526,
      "step": 8440
    },
    {
      "epoch": 0.755071039227951,
      "grad_norm": 0.3914361298084259,
      "learning_rate": 2e-05,
      "loss": 0.6921,
      "step": 8450
    },
    {
      "epoch": 0.7559646144223037,
      "grad_norm": 0.36308690905570984,
      "learning_rate": 2e-05,
      "loss": 0.6263,
      "step": 8460
    },
    {
      "epoch": 0.7568581896166563,
      "grad_norm": 0.3636039197444916,
      "learning_rate": 2e-05,
      "loss": 0.6371,
      "step": 8470
    },
    {
      "epoch": 0.7577517648110088,
      "grad_norm": 0.35106346011161804,
      "learning_rate": 2e-05,
      "loss": 0.6458,
      "step": 8480
    },
    {
      "epoch": 0.7586453400053614,
      "grad_norm": 0.3697260916233063,
      "learning_rate": 2e-05,
      "loss": 0.6416,
      "step": 8490
    },
    {
      "epoch": 0.7595389151997141,
      "grad_norm": 0.3724000155925751,
      "learning_rate": 2e-05,
      "loss": 0.662,
      "step": 8500
    },
    {
      "epoch": 0.7604324903940667,
      "grad_norm": 0.38306260108947754,
      "learning_rate": 2e-05,
      "loss": 0.6495,
      "step": 8510
    },
    {
      "epoch": 0.7613260655884193,
      "grad_norm": 0.40936288237571716,
      "learning_rate": 2e-05,
      "loss": 0.6538,
      "step": 8520
    },
    {
      "epoch": 0.7622196407827718,
      "grad_norm": 0.35177743434906006,
      "learning_rate": 2e-05,
      "loss": 0.6085,
      "step": 8530
    },
    {
      "epoch": 0.7631132159771244,
      "grad_norm": 0.37727177143096924,
      "learning_rate": 2e-05,
      "loss": 0.6587,
      "step": 8540
    },
    {
      "epoch": 0.7640067911714771,
      "grad_norm": 0.3676639795303345,
      "learning_rate": 2e-05,
      "loss": 0.6491,
      "step": 8550
    },
    {
      "epoch": 0.7649003663658297,
      "grad_norm": 0.3929283618927002,
      "learning_rate": 2e-05,
      "loss": 0.6448,
      "step": 8560
    },
    {
      "epoch": 0.7657939415601823,
      "grad_norm": 0.36641597747802734,
      "learning_rate": 2e-05,
      "loss": 0.6138,
      "step": 8570
    },
    {
      "epoch": 0.766687516754535,
      "grad_norm": 0.34161248803138733,
      "learning_rate": 2e-05,
      "loss": 0.6417,
      "step": 8580
    },
    {
      "epoch": 0.7675810919488875,
      "grad_norm": 0.3389277160167694,
      "learning_rate": 2e-05,
      "loss": 0.6214,
      "step": 8590
    },
    {
      "epoch": 0.7684746671432401,
      "grad_norm": 0.4783843159675598,
      "learning_rate": 2e-05,
      "loss": 0.6252,
      "step": 8600
    },
    {
      "epoch": 0.7693682423375927,
      "grad_norm": 0.39358949661254883,
      "learning_rate": 2e-05,
      "loss": 0.6304,
      "step": 8610
    },
    {
      "epoch": 0.7702618175319453,
      "grad_norm": 0.3984729051589966,
      "learning_rate": 2e-05,
      "loss": 0.6184,
      "step": 8620
    },
    {
      "epoch": 0.771155392726298,
      "grad_norm": 0.32481956481933594,
      "learning_rate": 2e-05,
      "loss": 0.6425,
      "step": 8630
    },
    {
      "epoch": 0.7720489679206505,
      "grad_norm": 0.33469006419181824,
      "learning_rate": 2e-05,
      "loss": 0.6217,
      "step": 8640
    },
    {
      "epoch": 0.7729425431150031,
      "grad_norm": 0.3726392686367035,
      "learning_rate": 2e-05,
      "loss": 0.6646,
      "step": 8650
    },
    {
      "epoch": 0.7738361183093557,
      "grad_norm": 0.3842551112174988,
      "learning_rate": 2e-05,
      "loss": 0.6722,
      "step": 8660
    },
    {
      "epoch": 0.7747296935037083,
      "grad_norm": 1.7092041969299316,
      "learning_rate": 2e-05,
      "loss": 0.6553,
      "step": 8670
    },
    {
      "epoch": 0.775623268698061,
      "grad_norm": 0.3700748085975647,
      "learning_rate": 2e-05,
      "loss": 0.6682,
      "step": 8680
    },
    {
      "epoch": 0.7765168438924136,
      "grad_norm": 0.3591102957725525,
      "learning_rate": 2e-05,
      "loss": 0.6483,
      "step": 8690
    },
    {
      "epoch": 0.7774104190867661,
      "grad_norm": 0.3569418489933014,
      "learning_rate": 2e-05,
      "loss": 0.6471,
      "step": 8700
    },
    {
      "epoch": 0.7783039942811187,
      "grad_norm": 0.434883713722229,
      "learning_rate": 2e-05,
      "loss": 0.6509,
      "step": 8710
    },
    {
      "epoch": 0.7791975694754714,
      "grad_norm": 0.3857283890247345,
      "learning_rate": 2e-05,
      "loss": 0.6429,
      "step": 8720
    },
    {
      "epoch": 0.780091144669824,
      "grad_norm": 0.37680357694625854,
      "learning_rate": 2e-05,
      "loss": 0.6624,
      "step": 8730
    },
    {
      "epoch": 0.7809847198641766,
      "grad_norm": 0.3501690626144409,
      "learning_rate": 2e-05,
      "loss": 0.6557,
      "step": 8740
    },
    {
      "epoch": 0.7818782950585291,
      "grad_norm": 0.37831151485443115,
      "learning_rate": 2e-05,
      "loss": 0.641,
      "step": 8750
    },
    {
      "epoch": 0.7827718702528818,
      "grad_norm": 0.3563501536846161,
      "learning_rate": 2e-05,
      "loss": 0.6135,
      "step": 8760
    },
    {
      "epoch": 0.7836654454472344,
      "grad_norm": 0.3351614475250244,
      "learning_rate": 2e-05,
      "loss": 0.6226,
      "step": 8770
    },
    {
      "epoch": 0.784559020641587,
      "grad_norm": 0.36878353357315063,
      "learning_rate": 2e-05,
      "loss": 0.6405,
      "step": 8780
    },
    {
      "epoch": 0.7854525958359396,
      "grad_norm": 0.33690643310546875,
      "learning_rate": 2e-05,
      "loss": 0.6178,
      "step": 8790
    },
    {
      "epoch": 0.7863461710302923,
      "grad_norm": 0.37017011642456055,
      "learning_rate": 2e-05,
      "loss": 0.653,
      "step": 8800
    },
    {
      "epoch": 0.7872397462246448,
      "grad_norm": 0.34322789311408997,
      "learning_rate": 2e-05,
      "loss": 0.6406,
      "step": 8810
    },
    {
      "epoch": 0.7881333214189974,
      "grad_norm": 0.34995952248573303,
      "learning_rate": 2e-05,
      "loss": 0.6013,
      "step": 8820
    },
    {
      "epoch": 0.78902689661335,
      "grad_norm": 0.37114983797073364,
      "learning_rate": 2e-05,
      "loss": 0.6235,
      "step": 8830
    },
    {
      "epoch": 0.7899204718077026,
      "grad_norm": 0.38480275869369507,
      "learning_rate": 2e-05,
      "loss": 0.6442,
      "step": 8840
    },
    {
      "epoch": 0.7908140470020553,
      "grad_norm": 0.38188016414642334,
      "learning_rate": 2e-05,
      "loss": 0.6256,
      "step": 8850
    },
    {
      "epoch": 0.7917076221964078,
      "grad_norm": 0.3496682643890381,
      "learning_rate": 2e-05,
      "loss": 0.6298,
      "step": 8860
    },
    {
      "epoch": 0.7926011973907604,
      "grad_norm": 0.4351358711719513,
      "learning_rate": 2e-05,
      "loss": 0.6273,
      "step": 8870
    },
    {
      "epoch": 0.793494772585113,
      "grad_norm": 0.34240037202835083,
      "learning_rate": 2e-05,
      "loss": 0.6234,
      "step": 8880
    },
    {
      "epoch": 0.7943883477794657,
      "grad_norm": 0.3621077239513397,
      "learning_rate": 2e-05,
      "loss": 0.6356,
      "step": 8890
    },
    {
      "epoch": 0.7952819229738183,
      "grad_norm": 0.3588218092918396,
      "learning_rate": 2e-05,
      "loss": 0.6144,
      "step": 8900
    },
    {
      "epoch": 0.7961754981681709,
      "grad_norm": 0.37342917919158936,
      "learning_rate": 2e-05,
      "loss": 0.6008,
      "step": 8910
    },
    {
      "epoch": 0.7970690733625234,
      "grad_norm": 0.33307990431785583,
      "learning_rate": 2e-05,
      "loss": 0.6543,
      "step": 8920
    },
    {
      "epoch": 0.797962648556876,
      "grad_norm": 0.351809024810791,
      "learning_rate": 2e-05,
      "loss": 0.6239,
      "step": 8930
    },
    {
      "epoch": 0.7988562237512287,
      "grad_norm": 0.3486623764038086,
      "learning_rate": 2e-05,
      "loss": 0.6359,
      "step": 8940
    },
    {
      "epoch": 0.7997497989455813,
      "grad_norm": 0.3517434597015381,
      "learning_rate": 2e-05,
      "loss": 0.6275,
      "step": 8950
    },
    {
      "epoch": 0.8006433741399339,
      "grad_norm": 0.3386448621749878,
      "learning_rate": 2e-05,
      "loss": 0.6235,
      "step": 8960
    },
    {
      "epoch": 0.8015369493342864,
      "grad_norm": 0.376968115568161,
      "learning_rate": 2e-05,
      "loss": 0.6534,
      "step": 8970
    },
    {
      "epoch": 0.802430524528639,
      "grad_norm": 0.34783679246902466,
      "learning_rate": 2e-05,
      "loss": 0.6293,
      "step": 8980
    },
    {
      "epoch": 0.8033240997229917,
      "grad_norm": 0.39423868060112,
      "learning_rate": 2e-05,
      "loss": 0.6388,
      "step": 8990
    },
    {
      "epoch": 0.8042176749173443,
      "grad_norm": 0.3771580755710602,
      "learning_rate": 2e-05,
      "loss": 0.6373,
      "step": 9000
    },
    {
      "epoch": 0.8051112501116969,
      "grad_norm": 0.3339296579360962,
      "learning_rate": 2e-05,
      "loss": 0.641,
      "step": 9010
    },
    {
      "epoch": 0.8060048253060496,
      "grad_norm": 0.3403621017932892,
      "learning_rate": 2e-05,
      "loss": 0.6528,
      "step": 9020
    },
    {
      "epoch": 0.8068984005004021,
      "grad_norm": 0.3941057622432709,
      "learning_rate": 2e-05,
      "loss": 0.6166,
      "step": 9030
    },
    {
      "epoch": 0.8077919756947547,
      "grad_norm": 0.38087400794029236,
      "learning_rate": 2e-05,
      "loss": 0.6492,
      "step": 9040
    },
    {
      "epoch": 0.8086855508891073,
      "grad_norm": 0.37871482968330383,
      "learning_rate": 2e-05,
      "loss": 0.6707,
      "step": 9050
    },
    {
      "epoch": 0.8095791260834599,
      "grad_norm": 0.35411202907562256,
      "learning_rate": 2e-05,
      "loss": 0.6133,
      "step": 9060
    },
    {
      "epoch": 0.8104727012778126,
      "grad_norm": 0.3310149013996124,
      "learning_rate": 2e-05,
      "loss": 0.6352,
      "step": 9070
    },
    {
      "epoch": 0.8113662764721651,
      "grad_norm": 0.3391226530075073,
      "learning_rate": 2e-05,
      "loss": 0.662,
      "step": 9080
    },
    {
      "epoch": 0.8122598516665177,
      "grad_norm": 0.3645947277545929,
      "learning_rate": 2e-05,
      "loss": 0.6707,
      "step": 9090
    },
    {
      "epoch": 0.8131534268608703,
      "grad_norm": 0.3577178418636322,
      "learning_rate": 2e-05,
      "loss": 0.6186,
      "step": 9100
    },
    {
      "epoch": 0.814047002055223,
      "grad_norm": 0.4456704556941986,
      "learning_rate": 2e-05,
      "loss": 0.6339,
      "step": 9110
    },
    {
      "epoch": 0.8149405772495756,
      "grad_norm": 0.3349698483943939,
      "learning_rate": 2e-05,
      "loss": 0.6225,
      "step": 9120
    },
    {
      "epoch": 0.8158341524439282,
      "grad_norm": 0.36627665162086487,
      "learning_rate": 2e-05,
      "loss": 0.6303,
      "step": 9130
    },
    {
      "epoch": 0.8167277276382807,
      "grad_norm": 0.3452802300453186,
      "learning_rate": 2e-05,
      "loss": 0.6563,
      "step": 9140
    },
    {
      "epoch": 0.8176213028326333,
      "grad_norm": 0.37613728642463684,
      "learning_rate": 2e-05,
      "loss": 0.6409,
      "step": 9150
    },
    {
      "epoch": 0.818514878026986,
      "grad_norm": 0.4117060899734497,
      "learning_rate": 2e-05,
      "loss": 0.6691,
      "step": 9160
    },
    {
      "epoch": 0.8194084532213386,
      "grad_norm": 0.38901233673095703,
      "learning_rate": 2e-05,
      "loss": 0.6158,
      "step": 9170
    },
    {
      "epoch": 0.8203020284156912,
      "grad_norm": 0.3740629255771637,
      "learning_rate": 2e-05,
      "loss": 0.602,
      "step": 9180
    },
    {
      "epoch": 0.8211956036100437,
      "grad_norm": 0.4940745532512665,
      "learning_rate": 2e-05,
      "loss": 0.6424,
      "step": 9190
    },
    {
      "epoch": 0.8220891788043964,
      "grad_norm": 0.3890405595302582,
      "learning_rate": 2e-05,
      "loss": 0.631,
      "step": 9200
    },
    {
      "epoch": 0.822982753998749,
      "grad_norm": 0.3665723204612732,
      "learning_rate": 2e-05,
      "loss": 0.6422,
      "step": 9210
    },
    {
      "epoch": 0.8238763291931016,
      "grad_norm": 0.36942023038864136,
      "learning_rate": 2e-05,
      "loss": 0.6371,
      "step": 9220
    },
    {
      "epoch": 0.8247699043874542,
      "grad_norm": 0.37345561385154724,
      "learning_rate": 2e-05,
      "loss": 0.6285,
      "step": 9230
    },
    {
      "epoch": 0.8256634795818069,
      "grad_norm": 0.4095766842365265,
      "learning_rate": 2e-05,
      "loss": 0.65,
      "step": 9240
    },
    {
      "epoch": 0.8265570547761594,
      "grad_norm": 0.3970649540424347,
      "learning_rate": 2e-05,
      "loss": 0.6435,
      "step": 9250
    },
    {
      "epoch": 0.827450629970512,
      "grad_norm": 0.37594643235206604,
      "learning_rate": 2e-05,
      "loss": 0.616,
      "step": 9260
    },
    {
      "epoch": 0.8283442051648646,
      "grad_norm": 0.35627245903015137,
      "learning_rate": 2e-05,
      "loss": 0.621,
      "step": 9270
    },
    {
      "epoch": 0.8292377803592172,
      "grad_norm": 0.38243454694747925,
      "learning_rate": 2e-05,
      "loss": 0.6447,
      "step": 9280
    },
    {
      "epoch": 0.8301313555535699,
      "grad_norm": 0.36381611227989197,
      "learning_rate": 2e-05,
      "loss": 0.6316,
      "step": 9290
    },
    {
      "epoch": 0.8310249307479224,
      "grad_norm": 0.32473254203796387,
      "learning_rate": 2e-05,
      "loss": 0.6306,
      "step": 9300
    },
    {
      "epoch": 0.831918505942275,
      "grad_norm": 0.37033504247665405,
      "learning_rate": 2e-05,
      "loss": 0.6533,
      "step": 9310
    },
    {
      "epoch": 0.8328120811366276,
      "grad_norm": 0.3761458098888397,
      "learning_rate": 2e-05,
      "loss": 0.6092,
      "step": 9320
    },
    {
      "epoch": 0.8337056563309803,
      "grad_norm": 0.3817290961742401,
      "learning_rate": 2e-05,
      "loss": 0.6333,
      "step": 9330
    },
    {
      "epoch": 0.8345992315253329,
      "grad_norm": 0.34578755497932434,
      "learning_rate": 2e-05,
      "loss": 0.6292,
      "step": 9340
    },
    {
      "epoch": 0.8354928067196855,
      "grad_norm": 0.3551556468009949,
      "learning_rate": 2e-05,
      "loss": 0.6045,
      "step": 9350
    },
    {
      "epoch": 0.836386381914038,
      "grad_norm": 0.3864268958568573,
      "learning_rate": 2e-05,
      "loss": 0.6265,
      "step": 9360
    },
    {
      "epoch": 0.8372799571083906,
      "grad_norm": 0.35280200839042664,
      "learning_rate": 2e-05,
      "loss": 0.6072,
      "step": 9370
    },
    {
      "epoch": 0.8381735323027433,
      "grad_norm": 0.6375495195388794,
      "learning_rate": 2e-05,
      "loss": 0.6224,
      "step": 9380
    },
    {
      "epoch": 0.8390671074970959,
      "grad_norm": 0.37343406677246094,
      "learning_rate": 2e-05,
      "loss": 0.6264,
      "step": 9390
    },
    {
      "epoch": 0.8399606826914485,
      "grad_norm": 0.35541436076164246,
      "learning_rate": 2e-05,
      "loss": 0.658,
      "step": 9400
    },
    {
      "epoch": 0.840854257885801,
      "grad_norm": 0.34526512026786804,
      "learning_rate": 2e-05,
      "loss": 0.6496,
      "step": 9410
    },
    {
      "epoch": 0.8417478330801537,
      "grad_norm": 0.39270293712615967,
      "learning_rate": 2e-05,
      "loss": 0.649,
      "step": 9420
    },
    {
      "epoch": 0.8426414082745063,
      "grad_norm": 0.3844796121120453,
      "learning_rate": 2e-05,
      "loss": 0.6209,
      "step": 9430
    },
    {
      "epoch": 0.8435349834688589,
      "grad_norm": 0.3906518220901489,
      "learning_rate": 2e-05,
      "loss": 0.6255,
      "step": 9440
    },
    {
      "epoch": 0.8444285586632115,
      "grad_norm": 0.3588649332523346,
      "learning_rate": 2e-05,
      "loss": 0.6174,
      "step": 9450
    },
    {
      "epoch": 0.8453221338575642,
      "grad_norm": 0.35434845089912415,
      "learning_rate": 2e-05,
      "loss": 0.645,
      "step": 9460
    },
    {
      "epoch": 0.8462157090519167,
      "grad_norm": 0.3549363613128662,
      "learning_rate": 2e-05,
      "loss": 0.6088,
      "step": 9470
    },
    {
      "epoch": 0.8471092842462693,
      "grad_norm": 0.3849292993545532,
      "learning_rate": 2e-05,
      "loss": 0.6323,
      "step": 9480
    },
    {
      "epoch": 0.8480028594406219,
      "grad_norm": 0.42119935154914856,
      "learning_rate": 2e-05,
      "loss": 0.6294,
      "step": 9490
    },
    {
      "epoch": 0.8488964346349746,
      "grad_norm": 0.3938457667827606,
      "learning_rate": 2e-05,
      "loss": 0.6469,
      "step": 9500
    },
    {
      "epoch": 0.8497900098293272,
      "grad_norm": 0.4140365421772003,
      "learning_rate": 2e-05,
      "loss": 0.6455,
      "step": 9510
    },
    {
      "epoch": 0.8506835850236797,
      "grad_norm": 0.33640751242637634,
      "learning_rate": 2e-05,
      "loss": 0.6123,
      "step": 9520
    },
    {
      "epoch": 0.8515771602180323,
      "grad_norm": 0.37525883316993713,
      "learning_rate": 2e-05,
      "loss": 0.605,
      "step": 9530
    },
    {
      "epoch": 0.8524707354123849,
      "grad_norm": 0.401439905166626,
      "learning_rate": 2e-05,
      "loss": 0.6503,
      "step": 9540
    },
    {
      "epoch": 0.8533643106067376,
      "grad_norm": 0.38618382811546326,
      "learning_rate": 2e-05,
      "loss": 0.6382,
      "step": 9550
    },
    {
      "epoch": 0.8542578858010902,
      "grad_norm": 0.35479140281677246,
      "learning_rate": 2e-05,
      "loss": 0.6268,
      "step": 9560
    },
    {
      "epoch": 0.8551514609954428,
      "grad_norm": 0.38308414816856384,
      "learning_rate": 2e-05,
      "loss": 0.6384,
      "step": 9570
    },
    {
      "epoch": 0.8560450361897953,
      "grad_norm": 0.3595040738582611,
      "learning_rate": 2e-05,
      "loss": 0.6276,
      "step": 9580
    },
    {
      "epoch": 0.856938611384148,
      "grad_norm": 0.34493914246559143,
      "learning_rate": 2e-05,
      "loss": 0.6463,
      "step": 9590
    },
    {
      "epoch": 0.8578321865785006,
      "grad_norm": 0.39991870522499084,
      "learning_rate": 2e-05,
      "loss": 0.6512,
      "step": 9600
    },
    {
      "epoch": 0.8587257617728532,
      "grad_norm": 0.37766894698143005,
      "learning_rate": 2e-05,
      "loss": 0.6505,
      "step": 9610
    },
    {
      "epoch": 0.8596193369672058,
      "grad_norm": 0.39890697598457336,
      "learning_rate": 2e-05,
      "loss": 0.6158,
      "step": 9620
    },
    {
      "epoch": 0.8605129121615583,
      "grad_norm": 0.379768967628479,
      "learning_rate": 2e-05,
      "loss": 0.6322,
      "step": 9630
    },
    {
      "epoch": 0.861406487355911,
      "grad_norm": 0.37261471152305603,
      "learning_rate": 2e-05,
      "loss": 0.6346,
      "step": 9640
    },
    {
      "epoch": 0.8623000625502636,
      "grad_norm": 0.4041343033313751,
      "learning_rate": 2e-05,
      "loss": 0.6336,
      "step": 9650
    },
    {
      "epoch": 0.8631936377446162,
      "grad_norm": 0.4248087406158447,
      "learning_rate": 2e-05,
      "loss": 0.6252,
      "step": 9660
    },
    {
      "epoch": 0.8640872129389688,
      "grad_norm": 0.3802678883075714,
      "learning_rate": 2e-05,
      "loss": 0.6302,
      "step": 9670
    },
    {
      "epoch": 0.8649807881333215,
      "grad_norm": 0.3724161684513092,
      "learning_rate": 2e-05,
      "loss": 0.6521,
      "step": 9680
    },
    {
      "epoch": 0.865874363327674,
      "grad_norm": 0.3731999397277832,
      "learning_rate": 2e-05,
      "loss": 0.6515,
      "step": 9690
    },
    {
      "epoch": 0.8667679385220266,
      "grad_norm": 0.3944038450717926,
      "learning_rate": 2e-05,
      "loss": 0.7015,
      "step": 9700
    },
    {
      "epoch": 0.8676615137163792,
      "grad_norm": 0.38460731506347656,
      "learning_rate": 2e-05,
      "loss": 0.6397,
      "step": 9710
    },
    {
      "epoch": 0.8685550889107319,
      "grad_norm": 0.36154255270957947,
      "learning_rate": 2e-05,
      "loss": 0.6379,
      "step": 9720
    },
    {
      "epoch": 0.8694486641050845,
      "grad_norm": 0.36715516448020935,
      "learning_rate": 2e-05,
      "loss": 0.6075,
      "step": 9730
    },
    {
      "epoch": 0.870342239299437,
      "grad_norm": 0.36314186453819275,
      "learning_rate": 2e-05,
      "loss": 0.6221,
      "step": 9740
    },
    {
      "epoch": 0.8712358144937896,
      "grad_norm": 0.39065611362457275,
      "learning_rate": 2e-05,
      "loss": 0.6179,
      "step": 9750
    },
    {
      "epoch": 0.8721293896881422,
      "grad_norm": 0.3411024212837219,
      "learning_rate": 2e-05,
      "loss": 0.6329,
      "step": 9760
    },
    {
      "epoch": 0.8730229648824949,
      "grad_norm": 0.3514428436756134,
      "learning_rate": 2e-05,
      "loss": 0.6563,
      "step": 9770
    },
    {
      "epoch": 0.8739165400768475,
      "grad_norm": 0.4481586217880249,
      "learning_rate": 2e-05,
      "loss": 0.6314,
      "step": 9780
    },
    {
      "epoch": 0.8748101152712001,
      "grad_norm": 0.3619888424873352,
      "learning_rate": 2e-05,
      "loss": 0.6629,
      "step": 9790
    },
    {
      "epoch": 0.8757036904655526,
      "grad_norm": 0.37259235978126526,
      "learning_rate": 2e-05,
      "loss": 0.6444,
      "step": 9800
    },
    {
      "epoch": 0.8765972656599053,
      "grad_norm": 0.4041177034378052,
      "learning_rate": 2e-05,
      "loss": 0.6447,
      "step": 9810
    },
    {
      "epoch": 0.8774908408542579,
      "grad_norm": 0.344304621219635,
      "learning_rate": 2e-05,
      "loss": 0.6256,
      "step": 9820
    },
    {
      "epoch": 0.8783844160486105,
      "grad_norm": 0.3712763786315918,
      "learning_rate": 2e-05,
      "loss": 0.5893,
      "step": 9830
    },
    {
      "epoch": 0.8792779912429631,
      "grad_norm": 0.5066938996315002,
      "learning_rate": 2e-05,
      "loss": 0.6407,
      "step": 9840
    },
    {
      "epoch": 0.8801715664373158,
      "grad_norm": 0.35774028301239014,
      "learning_rate": 2e-05,
      "loss": 0.6393,
      "step": 9850
    },
    {
      "epoch": 0.8810651416316683,
      "grad_norm": 0.39031165838241577,
      "learning_rate": 2e-05,
      "loss": 0.6428,
      "step": 9860
    },
    {
      "epoch": 0.8819587168260209,
      "grad_norm": 0.37468934059143066,
      "learning_rate": 2e-05,
      "loss": 0.6476,
      "step": 9870
    },
    {
      "epoch": 0.8828522920203735,
      "grad_norm": 0.33680132031440735,
      "learning_rate": 2e-05,
      "loss": 0.6382,
      "step": 9880
    },
    {
      "epoch": 0.8837458672147261,
      "grad_norm": 0.4165295660495758,
      "learning_rate": 2e-05,
      "loss": 0.6219,
      "step": 9890
    },
    {
      "epoch": 0.8846394424090788,
      "grad_norm": 0.35897859930992126,
      "learning_rate": 2e-05,
      "loss": 0.6376,
      "step": 9900
    },
    {
      "epoch": 0.8855330176034313,
      "grad_norm": 0.4254584312438965,
      "learning_rate": 2e-05,
      "loss": 0.6599,
      "step": 9910
    },
    {
      "epoch": 0.8864265927977839,
      "grad_norm": 0.3637544512748718,
      "learning_rate": 2e-05,
      "loss": 0.6163,
      "step": 9920
    },
    {
      "epoch": 0.8873201679921365,
      "grad_norm": 0.3738248348236084,
      "learning_rate": 2e-05,
      "loss": 0.638,
      "step": 9930
    },
    {
      "epoch": 0.8882137431864892,
      "grad_norm": 0.43496865034103394,
      "learning_rate": 2e-05,
      "loss": 0.635,
      "step": 9940
    },
    {
      "epoch": 0.8891073183808418,
      "grad_norm": 0.3840196132659912,
      "learning_rate": 2e-05,
      "loss": 0.6078,
      "step": 9950
    },
    {
      "epoch": 0.8900008935751944,
      "grad_norm": 0.35762760043144226,
      "learning_rate": 2e-05,
      "loss": 0.6153,
      "step": 9960
    },
    {
      "epoch": 0.8908944687695469,
      "grad_norm": 0.38396427035331726,
      "learning_rate": 2e-05,
      "loss": 0.66,
      "step": 9970
    },
    {
      "epoch": 0.8917880439638995,
      "grad_norm": 0.3843446969985962,
      "learning_rate": 2e-05,
      "loss": 0.6379,
      "step": 9980
    },
    {
      "epoch": 0.8926816191582522,
      "grad_norm": 0.39454856514930725,
      "learning_rate": 2e-05,
      "loss": 0.6298,
      "step": 9990
    },
    {
      "epoch": 0.8935751943526048,
      "grad_norm": 0.34030258655548096,
      "learning_rate": 2e-05,
      "loss": 0.6269,
      "step": 10000
    },
    {
      "epoch": 0.8944687695469574,
      "grad_norm": 0.366990864276886,
      "learning_rate": 2e-05,
      "loss": 0.615,
      "step": 10010
    },
    {
      "epoch": 0.8953623447413099,
      "grad_norm": 0.3749043345451355,
      "learning_rate": 2e-05,
      "loss": 0.651,
      "step": 10020
    },
    {
      "epoch": 0.8962559199356626,
      "grad_norm": 0.37012872099876404,
      "learning_rate": 2e-05,
      "loss": 0.6312,
      "step": 10030
    },
    {
      "epoch": 0.8971494951300152,
      "grad_norm": 0.354740709066391,
      "learning_rate": 2e-05,
      "loss": 0.6309,
      "step": 10040
    },
    {
      "epoch": 0.8980430703243678,
      "grad_norm": 0.3818489611148834,
      "learning_rate": 2e-05,
      "loss": 0.6269,
      "step": 10050
    },
    {
      "epoch": 0.8989366455187204,
      "grad_norm": 0.4574466049671173,
      "learning_rate": 2e-05,
      "loss": 0.6479,
      "step": 10060
    },
    {
      "epoch": 0.8998302207130731,
      "grad_norm": 0.3886832296848297,
      "learning_rate": 2e-05,
      "loss": 0.6453,
      "step": 10070
    },
    {
      "epoch": 0.9007237959074256,
      "grad_norm": 0.35603827238082886,
      "learning_rate": 2e-05,
      "loss": 0.6272,
      "step": 10080
    },
    {
      "epoch": 0.9016173711017782,
      "grad_norm": 0.39936068654060364,
      "learning_rate": 2e-05,
      "loss": 0.6073,
      "step": 10090
    },
    {
      "epoch": 0.9025109462961308,
      "grad_norm": 0.35282066464424133,
      "learning_rate": 2e-05,
      "loss": 0.6391,
      "step": 10100
    },
    {
      "epoch": 0.9034045214904834,
      "grad_norm": 0.6381559371948242,
      "learning_rate": 2e-05,
      "loss": 0.6122,
      "step": 10110
    },
    {
      "epoch": 0.9042980966848361,
      "grad_norm": 0.3749694228172302,
      "learning_rate": 2e-05,
      "loss": 0.6096,
      "step": 10120
    },
    {
      "epoch": 0.9051916718791886,
      "grad_norm": 0.384698748588562,
      "learning_rate": 2e-05,
      "loss": 0.6446,
      "step": 10130
    },
    {
      "epoch": 0.9060852470735412,
      "grad_norm": 0.38538843393325806,
      "learning_rate": 2e-05,
      "loss": 0.6115,
      "step": 10140
    },
    {
      "epoch": 0.9069788222678938,
      "grad_norm": 0.37189963459968567,
      "learning_rate": 2e-05,
      "loss": 0.6497,
      "step": 10150
    },
    {
      "epoch": 0.9078723974622465,
      "grad_norm": 0.35698720812797546,
      "learning_rate": 2e-05,
      "loss": 0.6677,
      "step": 10160
    },
    {
      "epoch": 0.9087659726565991,
      "grad_norm": 0.3853283226490021,
      "learning_rate": 2e-05,
      "loss": 0.653,
      "step": 10170
    },
    {
      "epoch": 0.9096595478509517,
      "grad_norm": 0.3594324290752411,
      "learning_rate": 1.999957358086795e-05,
      "loss": 0.6407,
      "step": 10180
    },
    {
      "epoch": 0.9105531230453042,
      "grad_norm": 0.34530362486839294,
      "learning_rate": 1.9991993813020483e-05,
      "loss": 0.6149,
      "step": 10190
    },
    {
      "epoch": 0.9114466982396568,
      "grad_norm": 0.36613914370536804,
      "learning_rate": 1.9974946373028628e-05,
      "loss": 0.6171,
      "step": 10200
    },
    {
      "epoch": 0.9123402734340095,
      "grad_norm": 0.36423757672309875,
      "learning_rate": 1.994844749501764e-05,
      "loss": 0.6376,
      "step": 10210
    },
    {
      "epoch": 0.9132338486283621,
      "grad_norm": 0.340941846370697,
      "learning_rate": 1.9912522413632988e-05,
      "loss": 0.631,
      "step": 10220
    },
    {
      "epoch": 0.9141274238227147,
      "grad_norm": 0.38077282905578613,
      "learning_rate": 1.9867205340009656e-05,
      "loss": 0.621,
      "step": 10230
    },
    {
      "epoch": 0.9150209990170672,
      "grad_norm": 0.34263554215431213,
      "learning_rate": 1.9812539429193164e-05,
      "loss": 0.614,
      "step": 10240
    },
    {
      "epoch": 0.9159145742114199,
      "grad_norm": 0.37161073088645935,
      "learning_rate": 1.9748576739043423e-05,
      "loss": 0.608,
      "step": 10250
    },
    {
      "epoch": 0.9168081494057725,
      "grad_norm": 0.3457585573196411,
      "learning_rate": 1.967537818066047e-05,
      "loss": 0.6334,
      "step": 10260
    },
    {
      "epoch": 0.9177017246001251,
      "grad_norm": 0.3803602159023285,
      "learning_rate": 1.959301346037941e-05,
      "loss": 0.6338,
      "step": 10270
    },
    {
      "epoch": 0.9185952997944777,
      "grad_norm": 0.3397793769836426,
      "learning_rate": 1.9501561013389646e-05,
      "loss": 0.6319,
      "step": 10280
    },
    {
      "epoch": 0.9194888749888304,
      "grad_norm": 0.3709755837917328,
      "learning_rate": 1.9401107929041795e-05,
      "loss": 0.6272,
      "step": 10290
    },
    {
      "epoch": 0.9203824501831829,
      "grad_norm": 0.39450544118881226,
      "learning_rate": 1.9291749867913225e-05,
      "loss": 0.6595,
      "step": 10300
    },
    {
      "epoch": 0.9212760253775355,
      "grad_norm": 0.3702409565448761,
      "learning_rate": 1.91735909707114e-05,
      "loss": 0.6397,
      "step": 10310
    },
    {
      "epoch": 0.9221696005718881,
      "grad_norm": 0.3482894003391266,
      "learning_rate": 1.9046743759101545e-05,
      "loss": 0.6119,
      "step": 10320
    },
    {
      "epoch": 0.9230631757662408,
      "grad_norm": 0.3624713122844696,
      "learning_rate": 1.8911329028553324e-05,
      "loss": 0.6297,
      "step": 10330
    },
    {
      "epoch": 0.9239567509605934,
      "grad_norm": 0.37890300154685974,
      "learning_rate": 1.876747573330835e-05,
      "loss": 0.6673,
      "step": 10340
    },
    {
      "epoch": 0.9248503261549459,
      "grad_norm": 0.33921119570732117,
      "learning_rate": 1.8615320863578244e-05,
      "loss": 0.6199,
      "step": 10350
    },
    {
      "epoch": 0.9257439013492985,
      "grad_norm": 0.39643770456314087,
      "learning_rate": 1.8455009315090036e-05,
      "loss": 0.6311,
      "step": 10360
    },
    {
      "epoch": 0.9266374765436511,
      "grad_norm": 0.35413259267807007,
      "learning_rate": 1.8286693751103294e-05,
      "loss": 0.6443,
      "step": 10370
    },
    {
      "epoch": 0.9275310517380038,
      "grad_norm": 0.5789254307746887,
      "learning_rate": 1.811053445703024e-05,
      "loss": 0.6407,
      "step": 10380
    },
    {
      "epoch": 0.9284246269323564,
      "grad_norm": 0.44033607840538025,
      "learning_rate": 1.7926699187797375e-05,
      "loss": 0.6197,
      "step": 10390
    },
    {
      "epoch": 0.929318202126709,
      "grad_norm": 0.3453246057033539,
      "learning_rate": 1.7735363008093995e-05,
      "loss": 0.6211,
      "step": 10400
    },
    {
      "epoch": 0.9302117773210615,
      "grad_norm": 0.38900038599967957,
      "learning_rate": 1.7536708125659577e-05,
      "loss": 0.6117,
      "step": 10410
    },
    {
      "epoch": 0.9311053525154142,
      "grad_norm": 0.3746323883533478,
      "learning_rate": 1.733092371776905e-05,
      "loss": 0.6197,
      "step": 10420
    },
    {
      "epoch": 0.9319989277097668,
      "grad_norm": 0.35671037435531616,
      "learning_rate": 1.7118205751080915e-05,
      "loss": 0.6307,
      "step": 10430
    },
    {
      "epoch": 0.9328925029041194,
      "grad_norm": 0.3590388298034668,
      "learning_rate": 1.6898756795019935e-05,
      "loss": 0.6527,
      "step": 10440
    },
    {
      "epoch": 0.933786078098472,
      "grad_norm": 0.38040000200271606,
      "learning_rate": 1.667278582887211e-05,
      "loss": 0.6269,
      "step": 10450
    },
    {
      "epoch": 0.9346796532928245,
      "grad_norm": 0.3766019344329834,
      "learning_rate": 1.6440508042775545e-05,
      "loss": 0.6014,
      "step": 10460
    },
    {
      "epoch": 0.9355732284871772,
      "grad_norm": 0.40499037504196167,
      "learning_rate": 1.6202144632796796e-05,
      "loss": 0.6353,
      "step": 10470
    },
    {
      "epoch": 0.9364668036815298,
      "grad_norm": 0.3391999304294586,
      "learning_rate": 1.595792259028783e-05,
      "loss": 0.6451,
      "step": 10480
    },
    {
      "epoch": 0.9373603788758824,
      "grad_norm": 0.3738960921764374,
      "learning_rate": 1.570807448572419e-05,
      "loss": 0.6377,
      "step": 10490
    },
    {
      "epoch": 0.938253954070235,
      "grad_norm": 0.3816886246204376,
      "learning_rate": 1.5452838247230197e-05,
      "loss": 0.5896,
      "step": 10500
    },
    {
      "epoch": 0.9391475292645877,
      "grad_norm": 0.3897479176521301,
      "learning_rate": 1.5192456934002106e-05,
      "loss": 0.6202,
      "step": 10510
    },
    {
      "epoch": 0.9400411044589402,
      "grad_norm": 0.3505251407623291,
      "learning_rate": 1.492717850484502e-05,
      "loss": 0.6434,
      "step": 10520
    },
    {
      "epoch": 0.9409346796532928,
      "grad_norm": 0.35288867354393005,
      "learning_rate": 1.4657255582043904e-05,
      "loss": 0.6283,
      "step": 10530
    },
    {
      "epoch": 0.9418282548476454,
      "grad_norm": 0.3790437877178192,
      "learning_rate": 1.4382945210793663e-05,
      "loss": 0.6203,
      "step": 10540
    },
    {
      "epoch": 0.942721830041998,
      "grad_norm": 0.3956012725830078,
      "learning_rate": 1.4104508614417267e-05,
      "loss": 0.6169,
      "step": 10550
    },
    {
      "epoch": 0.9436154052363507,
      "grad_norm": 0.37420955300331116,
      "learning_rate": 1.382221094560512e-05,
      "loss": 0.6093,
      "step": 10560
    },
    {
      "epoch": 0.9445089804307032,
      "grad_norm": 0.3559351861476898,
      "learning_rate": 1.3536321033912506e-05,
      "loss": 0.6167,
      "step": 10570
    },
    {
      "epoch": 0.9454025556250558,
      "grad_norm": 0.36077457666397095,
      "learning_rate": 1.3247111129755609e-05,
      "loss": 0.6366,
      "step": 10580
    },
    {
      "epoch": 0.9462961308194084,
      "grad_norm": 0.5243368148803711,
      "learning_rate": 1.2954856645149868e-05,
      "loss": 0.6255,
      "step": 10590
    },
    {
      "epoch": 0.9471897060137611,
      "grad_norm": 0.3909355103969574,
      "learning_rate": 1.2659835891437553e-05,
      "loss": 0.6264,
      "step": 10600
    },
    {
      "epoch": 0.9480832812081137,
      "grad_norm": 0.39675119519233704,
      "learning_rate": 1.236232981425436e-05,
      "loss": 0.6506,
      "step": 10610
    },
    {
      "epoch": 0.9489768564024663,
      "grad_norm": 0.3684333264827728,
      "learning_rate": 1.2062621725987402e-05,
      "loss": 0.6333,
      "step": 10620
    },
    {
      "epoch": 0.9498704315968188,
      "grad_norm": 0.3695078194141388,
      "learning_rate": 1.1760997035979318e-05,
      "loss": 0.6108,
      "step": 10630
    },
    {
      "epoch": 0.9507640067911715,
      "grad_norm": 0.32515373826026917,
      "learning_rate": 1.1457742978735514e-05,
      "loss": 0.6411,
      "step": 10640
    },
    {
      "epoch": 0.9516575819855241,
      "grad_norm": 0.3982977271080017,
      "learning_rate": 1.1153148340393298e-05,
      "loss": 0.651,
      "step": 10650
    },
    {
      "epoch": 0.9525511571798767,
      "grad_norm": 0.35070690512657166,
      "learning_rate": 1.0847503183713424e-05,
      "loss": 0.5966,
      "step": 10660
    },
    {
      "epoch": 0.9534447323742293,
      "grad_norm": 0.3726920485496521,
      "learning_rate": 1.0541098571855896e-05,
      "loss": 0.5914,
      "step": 10670
    },
    {
      "epoch": 0.9543383075685818,
      "grad_norm": 0.3748769462108612,
      "learning_rate": 1.0234226291203123e-05,
      "loss": 0.6023,
      "step": 10680
    },
    {
      "epoch": 0.9552318827629345,
      "grad_norm": 0.3589731752872467,
      "learning_rate": 9.927178573494333e-06,
      "loss": 0.6206,
      "step": 10690
    },
    {
      "epoch": 0.9561254579572871,
      "grad_norm": 0.4007958471775055,
      "learning_rate": 9.620247817535902e-06,
      "loss": 0.632,
      "step": 10700
    },
    {
      "epoch": 0.9570190331516397,
      "grad_norm": 0.3885563910007477,
      "learning_rate": 9.313726310752581e-06,
      "loss": 0.6009,
      "step": 10710
    },
    {
      "epoch": 0.9579126083459923,
      "grad_norm": 0.3295983076095581,
      "learning_rate": 9.00790595084477e-06,
      "loss": 0.5876,
      "step": 10720
    },
    {
      "epoch": 0.958806183540345,
      "grad_norm": 0.37213143706321716,
      "learning_rate": 8.703077967816944e-06,
      "loss": 0.6344,
      "step": 10730
    },
    {
      "epoch": 0.9596997587346975,
      "grad_norm": 0.3792665898799896,
      "learning_rate": 8.39953264664192e-06,
      "loss": 0.6265,
      "step": 10740
    },
    {
      "epoch": 0.9605933339290501,
      "grad_norm": 0.3770064413547516,
      "learning_rate": 8.097559050825058e-06,
      "loss": 0.5898,
      "step": 10750
    },
    {
      "epoch": 0.9614869091234027,
      "grad_norm": 0.3506534993648529,
      "learning_rate": 7.79744474713166e-06,
      "loss": 0.5981,
      "step": 10760
    },
    {
      "epoch": 0.9623804843177554,
      "grad_norm": 0.3346441686153412,
      "learning_rate": 7.499475531739696e-06,
      "loss": 0.6016,
      "step": 10770
    },
    {
      "epoch": 0.963274059512108,
      "grad_norm": 0.36221516132354736,
      "learning_rate": 7.203935158078662e-06,
      "loss": 0.6235,
      "step": 10780
    },
    {
      "epoch": 0.9641676347064605,
      "grad_norm": 0.3948242664337158,
      "learning_rate": 6.911105066613709e-06,
      "loss": 0.6125,
      "step": 10790
    },
    {
      "epoch": 0.9650612099008131,
      "grad_norm": 0.36343660950660706,
      "learning_rate": 6.621264116832394e-06,
      "loss": 0.6146,
      "step": 10800
    },
    {
      "epoch": 0.9659547850951657,
      "grad_norm": 0.35896793007850647,
      "learning_rate": 6.334688321689274e-06,
      "loss": 0.6089,
      "step": 10810
    },
    {
      "epoch": 0.9668483602895184,
      "grad_norm": 0.339715838432312,
      "learning_rate": 6.051650584761237e-06,
      "loss": 0.5874,
      "step": 10820
    },
    {
      "epoch": 0.967741935483871,
      "grad_norm": 0.40527647733688354,
      "learning_rate": 5.772420440363841e-06,
      "loss": 0.6445,
      "step": 10830
    },
    {
      "epoch": 0.9686355106782236,
      "grad_norm": 0.3736339807510376,
      "learning_rate": 5.497263796876207e-06,
      "loss": 0.6156,
      "step": 10840
    },
    {
      "epoch": 0.9695290858725761,
      "grad_norm": 0.35081949830055237,
      "learning_rate": 5.226442683518813e-06,
      "loss": 0.5954,
      "step": 10850
    },
    {
      "epoch": 0.9704226610669288,
      "grad_norm": 0.3600223958492279,
      "learning_rate": 4.960215000825432e-06,
      "loss": 0.6168,
      "step": 10860
    },
    {
      "epoch": 0.9713162362612814,
      "grad_norm": 0.3501090705394745,
      "learning_rate": 4.698834275046711e-06,
      "loss": 0.6088,
      "step": 10870
    },
    {
      "epoch": 0.972209811455634,
      "grad_norm": 0.4176599383354187,
      "learning_rate": 4.442549416719423e-06,
      "loss": 0.6141,
      "step": 10880
    },
    {
      "epoch": 0.9731033866499866,
      "grad_norm": 0.3563132584095001,
      "learning_rate": 4.1916044836311565e-06,
      "loss": 0.5642,
      "step": 10890
    },
    {
      "epoch": 0.9739969618443391,
      "grad_norm": 0.39151182770729065,
      "learning_rate": 3.94623844840626e-06,
      "loss": 0.6153,
      "step": 10900
    },
    {
      "epoch": 0.9748905370386918,
      "grad_norm": 0.39762741327285767,
      "learning_rate": 3.7066849709343467e-06,
      "loss": 0.6076,
      "step": 10910
    },
    {
      "epoch": 0.9757841122330444,
      "grad_norm": 0.3709637522697449,
      "learning_rate": 3.473172175858059e-06,
      "loss": 0.6102,
      "step": 10920
    },
    {
      "epoch": 0.976677687427397,
      "grad_norm": 0.36330679059028625,
      "learning_rate": 3.2459224353320046e-06,
      "loss": 0.6097,
      "step": 10930
    },
    {
      "epoch": 0.9775712626217496,
      "grad_norm": 0.4118615388870239,
      "learning_rate": 3.025152157259735e-06,
      "loss": 0.6139,
      "step": 10940
    },
    {
      "epoch": 0.9784648378161023,
      "grad_norm": 0.3405228555202484,
      "learning_rate": 2.811071579210425e-06,
      "loss": 0.6373,
      "step": 10950
    },
    {
      "epoch": 0.9793584130104548,
      "grad_norm": 0.3809181749820709,
      "learning_rate": 2.603884568211513e-06,
      "loss": 0.6344,
      "step": 10960
    },
    {
      "epoch": 0.9802519882048074,
      "grad_norm": 0.42268434166908264,
      "learning_rate": 2.4037884266079434e-06,
      "loss": 0.6208,
      "step": 10970
    },
    {
      "epoch": 0.98114556339916,
      "grad_norm": 0.3646038770675659,
      "learning_rate": 2.2109737041728955e-06,
      "loss": 0.6321,
      "step": 10980
    },
    {
      "epoch": 0.9820391385935127,
      "grad_norm": 0.3551325500011444,
      "learning_rate": 2.0256240166489454e-06,
      "loss": 0.5831,
      "step": 10990
    },
    {
      "epoch": 0.9829327137878653,
      "grad_norm": 0.3633440136909485,
      "learning_rate": 1.847915870892413e-06,
      "loss": 0.6153,
      "step": 11000
    },
    {
      "epoch": 0.9838262889822178,
      "grad_norm": 0.34125491976737976,
      "learning_rate": 1.6780184967874715e-06,
      "loss": 0.5976,
      "step": 11010
    },
    {
      "epoch": 0.9847198641765704,
      "grad_norm": 0.3639891445636749,
      "learning_rate": 1.5160936860900294e-06,
      "loss": 0.6065,
      "step": 11020
    },
    {
      "epoch": 0.985613439370923,
      "grad_norm": 0.36485856771469116,
      "learning_rate": 1.3622956383548762e-06,
      "loss": 0.5991,
      "step": 11030
    },
    {
      "epoch": 0.9865070145652757,
      "grad_norm": 0.35530680418014526,
      "learning_rate": 1.216770814092834e-06,
      "loss": 0.6006,
      "step": 11040
    },
    {
      "epoch": 0.9874005897596283,
      "grad_norm": 0.373453289270401,
      "learning_rate": 1.0796577952977013e-06,
      "loss": 0.5755,
      "step": 11050
    },
    {
      "epoch": 0.9882941649539809,
      "grad_norm": 0.3445267379283905,
      "learning_rate": 9.510871534758705e-07,
      "loss": 0.5787,
      "step": 11060
    },
    {
      "epoch": 0.9891877401483334,
      "grad_norm": 0.4402855634689331,
      "learning_rate": 8.311813253042334e-07,
      "loss": 0.6039,
      "step": 11070
    },
    {
      "epoch": 0.9900813153426861,
      "grad_norm": 0.35663896799087524,
      "learning_rate": 7.200544960348308e-07,
      "loss": 0.6333,
      "step": 11080
    },
    {
      "epoch": 0.9909748905370387,
      "grad_norm": 0.4121132791042328,
      "learning_rate": 6.178124907572545e-07,
      "loss": 0.579,
      "step": 11090
    },
    {
      "epoch": 0.9918684657313913,
      "grad_norm": 0.44549524784088135,
      "learning_rate": 5.245526736223528e-07,
      "loss": 0.6012,
      "step": 11100
    },
    {
      "epoch": 0.9927620409257439,
      "grad_norm": 0.35994240641593933,
      "learning_rate": 4.403638551232205e-07,
      "loss": 0.6002,
      "step": 11110
    },
    {
      "epoch": 0.9936556161200965,
      "grad_norm": 0.37222903966903687,
      "learning_rate": 3.6532620752175616e-07,
      "loss": 0.5965,
      "step": 11120
    },
    {
      "epoch": 0.9945491913144491,
      "grad_norm": 0.3795420825481415,
      "learning_rate": 2.9951118850133247e-07,
      "loss": 0.5871,
      "step": 11130
    },
    {
      "epoch": 0.9954427665088017,
      "grad_norm": 0.3823469281196594,
      "learning_rate": 2.429814731182774e-07,
      "loss": 0.6353,
      "step": 11140
    },
    {
      "epoch": 0.9963363417031543,
      "grad_norm": 0.37609076499938965,
      "learning_rate": 1.957908941169813e-07,
      "loss": 0.6181,
      "step": 11150
    },
    {
      "epoch": 0.997229916897507,
      "grad_norm": 0.40547484159469604,
      "learning_rate": 1.5798439066545254e-07,
      "loss": 0.5868,
      "step": 11160
    },
    {
      "epoch": 0.9981234920918596,
      "grad_norm": 0.3694177567958832,
      "learning_rate": 1.2959796556014471e-07,
      "loss": 0.6237,
      "step": 11170
    },
    {
      "epoch": 0.9990170672862121,
      "grad_norm": 0.3703685998916626,
      "learning_rate": 1.1065865094081884e-07,
      "loss": 0.5878,
      "step": 11180
    },
    {
      "epoch": 0.9999106424805647,
      "grad_norm": 0.3740469813346863,
      "learning_rate": 1.011844825480734e-07,
      "loss": 0.6155,
      "step": 11190
    },
    {
      "epoch": 1.0002680725583057,
      "step": 11194,
      "total_flos": 4.547514943786923e+19,
      "train_loss": 0.0034757169076368513,
      "train_runtime": 610.7283,
      "train_samples_per_second": 219.947,
      "train_steps_per_second": 18.329
    }
  ],
  "logging_steps": 10,
  "max_steps": 11194,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 159,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.547514943786923e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}