{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9837405519423449,
  "eval_steps": 500,
  "global_step": 11193,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008788890841975743,
      "grad_norm": 0.7299243807792664,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 0.5786,
      "step": 10
    },
    {
      "epoch": 0.0017577781683951485,
      "grad_norm": 0.7039077281951904,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 0.5844,
      "step": 20
    },
    {
      "epoch": 0.002636667252592723,
      "grad_norm": 0.594024121761322,
      "learning_rate": 3.342134831460675e-06,
      "loss": 0.5971,
      "step": 30
    },
    {
      "epoch": 0.003515556336790297,
      "grad_norm": 0.7417164444923401,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 0.5748,
      "step": 40
    },
    {
      "epoch": 0.004394445420987872,
      "grad_norm": 0.9814750552177429,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 0.5991,
      "step": 50
    },
    {
      "epoch": 0.005273334505185446,
      "grad_norm": 0.6245082020759583,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 0.5847,
      "step": 60
    },
    {
      "epoch": 0.00615222358938302,
      "grad_norm": 0.5894784331321716,
      "learning_rate": 7.814044943820226e-06,
      "loss": 0.5843,
      "step": 70
    },
    {
      "epoch": 0.007031112673580594,
      "grad_norm": 0.6612409353256226,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.572,
      "step": 80
    },
    {
      "epoch": 0.007910001757778168,
      "grad_norm": 0.614031195640564,
      "learning_rate": 1.005e-05,
      "loss": 0.5755,
      "step": 90
    },
    {
      "epoch": 0.008788890841975743,
      "grad_norm": 0.6163526773452759,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.5672,
      "step": 100
    },
    {
      "epoch": 0.009667779926173316,
      "grad_norm": 0.6364962458610535,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.5639,
      "step": 110
    },
    {
      "epoch": 0.010546669010370891,
      "grad_norm": 0.6238141655921936,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.5805,
      "step": 120
    },
    {
      "epoch": 0.011425558094568465,
      "grad_norm": 0.7033671736717224,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.5794,
      "step": 130
    },
    {
      "epoch": 0.01230444717876604,
      "grad_norm": 0.5810057520866394,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.5318,
      "step": 140
    },
    {
      "epoch": 0.013183336262963615,
      "grad_norm": 0.6278232336044312,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.5694,
      "step": 150
    },
    {
      "epoch": 0.014062225347161188,
      "grad_norm": 0.5975486040115356,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.5692,
      "step": 160
    },
    {
      "epoch": 0.014941114431358763,
      "grad_norm": 0.582416296005249,
      "learning_rate": 1.89938202247191e-05,
      "loss": 0.5773,
      "step": 170
    },
    {
      "epoch": 0.015820003515556336,
      "grad_norm": 0.6315813660621643,
      "learning_rate": 2e-05,
      "loss": 0.5881,
      "step": 180
    },
    {
      "epoch": 0.01669889259975391,
      "grad_norm": 0.5048695802688599,
      "learning_rate": 2e-05,
      "loss": 0.6048,
      "step": 190
    },
    {
      "epoch": 0.017577781683951486,
      "grad_norm": 0.5755158066749573,
      "learning_rate": 2e-05,
      "loss": 0.5961,
      "step": 200
    },
    {
      "epoch": 0.01845667076814906,
      "grad_norm": 0.46752920746803284,
      "learning_rate": 2e-05,
      "loss": 0.5767,
      "step": 210
    },
    {
      "epoch": 0.019335559852346633,
      "grad_norm": 0.5730629563331604,
      "learning_rate": 2e-05,
      "loss": 0.593,
      "step": 220
    },
    {
      "epoch": 0.02021444893654421,
      "grad_norm": 0.4156428575515747,
      "learning_rate": 2e-05,
      "loss": 0.5804,
      "step": 230
    },
    {
      "epoch": 0.021093338020741783,
      "grad_norm": 0.4448103606700897,
      "learning_rate": 2e-05,
      "loss": 0.5891,
      "step": 240
    },
    {
      "epoch": 0.021972227104939356,
      "grad_norm": 0.446362167596817,
      "learning_rate": 2e-05,
      "loss": 0.5669,
      "step": 250
    },
    {
      "epoch": 0.02285111618913693,
      "grad_norm": 0.42595115303993225,
      "learning_rate": 2e-05,
      "loss": 0.5885,
      "step": 260
    },
    {
      "epoch": 0.023730005273334506,
      "grad_norm": 0.45942461490631104,
      "learning_rate": 2e-05,
      "loss": 0.5812,
      "step": 270
    },
    {
      "epoch": 0.02460889435753208,
      "grad_norm": 0.4314509332180023,
      "learning_rate": 2e-05,
      "loss": 0.5984,
      "step": 280
    },
    {
      "epoch": 0.025487783441729653,
      "grad_norm": 0.4355176091194153,
      "learning_rate": 2e-05,
      "loss": 0.5995,
      "step": 290
    },
    {
      "epoch": 0.02636667252592723,
      "grad_norm": 0.4335140287876129,
      "learning_rate": 2e-05,
      "loss": 0.5725,
      "step": 300
    },
    {
      "epoch": 0.027245561610124803,
      "grad_norm": 0.41898155212402344,
      "learning_rate": 2e-05,
      "loss": 0.5947,
      "step": 310
    },
    {
      "epoch": 0.028124450694322376,
      "grad_norm": 0.35632404685020447,
      "learning_rate": 2e-05,
      "loss": 0.5575,
      "step": 320
    },
    {
      "epoch": 0.02900333977851995,
      "grad_norm": 0.4908845126628876,
      "learning_rate": 2e-05,
      "loss": 0.5874,
      "step": 330
    },
    {
      "epoch": 0.029882228862717526,
      "grad_norm": 0.39167770743370056,
      "learning_rate": 2e-05,
      "loss": 0.6033,
      "step": 340
    },
    {
      "epoch": 0.0307611179469151,
      "grad_norm": 0.41944095492362976,
      "learning_rate": 2e-05,
      "loss": 0.578,
      "step": 350
    },
    {
      "epoch": 0.03164000703111267,
      "grad_norm": 0.4073144197463989,
      "learning_rate": 2e-05,
      "loss": 0.5851,
      "step": 360
    },
    {
      "epoch": 0.032518896115310246,
      "grad_norm": 0.3751962184906006,
      "learning_rate": 2e-05,
      "loss": 0.5799,
      "step": 370
    },
    {
      "epoch": 0.03339778519950782,
      "grad_norm": 0.39586400985717773,
      "learning_rate": 2e-05,
      "loss": 0.5798,
      "step": 380
    },
    {
      "epoch": 0.0342766742837054,
      "grad_norm": 0.4123347997665405,
      "learning_rate": 2e-05,
      "loss": 0.574,
      "step": 390
    },
    {
      "epoch": 0.03515556336790297,
      "grad_norm": 0.4077567160129547,
      "learning_rate": 2e-05,
      "loss": 0.6083,
      "step": 400
    },
    {
      "epoch": 0.036034452452100546,
      "grad_norm": 0.36184021830558777,
      "learning_rate": 2e-05,
      "loss": 0.5791,
      "step": 410
    },
    {
      "epoch": 0.03691334153629812,
      "grad_norm": 0.39715006947517395,
      "learning_rate": 2e-05,
      "loss": 0.5809,
      "step": 420
    },
    {
      "epoch": 0.03779223062049569,
      "grad_norm": 0.34982648491859436,
      "learning_rate": 2e-05,
      "loss": 0.5651,
      "step": 430
    },
    {
      "epoch": 0.038671119704693266,
      "grad_norm": 0.35893237590789795,
      "learning_rate": 2e-05,
      "loss": 0.581,
      "step": 440
    },
    {
      "epoch": 0.03955000878889084,
      "grad_norm": 0.38406050205230713,
      "learning_rate": 2e-05,
      "loss": 0.5961,
      "step": 450
    },
    {
      "epoch": 0.04042889787308842,
      "grad_norm": 0.4047548174858093,
      "learning_rate": 2e-05,
      "loss": 0.561,
      "step": 460
    },
    {
      "epoch": 0.04130778695728599,
      "grad_norm": 0.3789917826652527,
      "learning_rate": 2e-05,
      "loss": 0.5899,
      "step": 470
    },
    {
      "epoch": 0.042186676041483566,
      "grad_norm": 0.40871259570121765,
      "learning_rate": 2e-05,
      "loss": 0.5848,
      "step": 480
    },
    {
      "epoch": 0.04306556512568114,
      "grad_norm": 0.3634383976459503,
      "learning_rate": 2e-05,
      "loss": 0.5942,
      "step": 490
    },
    {
      "epoch": 0.04394445420987871,
      "grad_norm": 0.3622599244117737,
      "learning_rate": 2e-05,
      "loss": 0.5632,
      "step": 500
    },
    {
      "epoch": 0.044823343294076286,
      "grad_norm": 0.3681706488132477,
      "learning_rate": 2e-05,
      "loss": 0.5918,
      "step": 510
    },
    {
      "epoch": 0.04570223237827386,
      "grad_norm": 0.4069708287715912,
      "learning_rate": 2e-05,
      "loss": 0.5745,
      "step": 520
    },
    {
      "epoch": 0.04658112146247144,
      "grad_norm": 0.40182244777679443,
      "learning_rate": 2e-05,
      "loss": 0.575,
      "step": 530
    },
    {
      "epoch": 0.04746001054666901,
      "grad_norm": 0.39383378624916077,
      "learning_rate": 2e-05,
      "loss": 0.5431,
      "step": 540
    },
    {
      "epoch": 0.048338899630866586,
      "grad_norm": 0.35547786951065063,
      "learning_rate": 2e-05,
      "loss": 0.5653,
      "step": 550
    },
    {
      "epoch": 0.04921778871506416,
      "grad_norm": 0.35708558559417725,
      "learning_rate": 2e-05,
      "loss": 0.5726,
      "step": 560
    },
    {
      "epoch": 0.05009667779926173,
      "grad_norm": 0.41257601976394653,
      "learning_rate": 2e-05,
      "loss": 0.5546,
      "step": 570
    },
    {
      "epoch": 0.050975566883459306,
      "grad_norm": 0.37748679518699646,
      "learning_rate": 2e-05,
      "loss": 0.5796,
      "step": 580
    },
    {
      "epoch": 0.05185445596765688,
      "grad_norm": 0.3628121614456177,
      "learning_rate": 2e-05,
      "loss": 0.5542,
      "step": 590
    },
    {
      "epoch": 0.05273334505185446,
      "grad_norm": 0.31530919671058655,
      "learning_rate": 2e-05,
      "loss": 0.5643,
      "step": 600
    },
    {
      "epoch": 0.05361223413605203,
      "grad_norm": 0.3966090679168701,
      "learning_rate": 2e-05,
      "loss": 0.588,
      "step": 610
    },
    {
      "epoch": 0.054491123220249606,
      "grad_norm": 0.41289761662483215,
      "learning_rate": 2e-05,
      "loss": 0.5868,
      "step": 620
    },
    {
      "epoch": 0.05537001230444718,
      "grad_norm": 0.37477827072143555,
      "learning_rate": 2e-05,
      "loss": 0.5764,
      "step": 630
    },
    {
      "epoch": 0.05624890138864475,
      "grad_norm": 0.4318097233772278,
      "learning_rate": 2e-05,
      "loss": 0.581,
      "step": 640
    },
    {
      "epoch": 0.057127790472842326,
      "grad_norm": 0.36890605092048645,
      "learning_rate": 2e-05,
      "loss": 0.5623,
      "step": 650
    },
    {
      "epoch": 0.0580066795570399,
      "grad_norm": 0.409492164850235,
      "learning_rate": 2e-05,
      "loss": 0.5848,
      "step": 660
    },
    {
      "epoch": 0.05888556864123748,
      "grad_norm": 0.3958004415035248,
      "learning_rate": 2e-05,
      "loss": 0.5851,
      "step": 670
    },
    {
      "epoch": 0.05976445772543505,
      "grad_norm": 0.3854322135448456,
      "learning_rate": 2e-05,
      "loss": 0.5867,
      "step": 680
    },
    {
      "epoch": 0.060643346809632626,
      "grad_norm": 0.3919129967689514,
      "learning_rate": 2e-05,
      "loss": 0.5859,
      "step": 690
    },
    {
      "epoch": 0.0615222358938302,
      "grad_norm": 0.34421873092651367,
      "learning_rate": 2e-05,
      "loss": 0.5695,
      "step": 700
    },
    {
      "epoch": 0.06240112497802777,
      "grad_norm": 0.36081117391586304,
      "learning_rate": 2e-05,
      "loss": 0.561,
      "step": 710
    },
    {
      "epoch": 0.06328001406222535,
      "grad_norm": 0.3740138113498688,
      "learning_rate": 2e-05,
      "loss": 0.5705,
      "step": 720
    },
    {
      "epoch": 0.06415890314642292,
      "grad_norm": 0.35506170988082886,
      "learning_rate": 2e-05,
      "loss": 0.5534,
      "step": 730
    },
    {
      "epoch": 0.06503779223062049,
      "grad_norm": 0.380117267370224,
      "learning_rate": 2e-05,
      "loss": 0.5559,
      "step": 740
    },
    {
      "epoch": 0.06591668131481807,
      "grad_norm": 0.36159446835517883,
      "learning_rate": 2e-05,
      "loss": 0.5886,
      "step": 750
    },
    {
      "epoch": 0.06679557039901564,
      "grad_norm": 0.3657621145248413,
      "learning_rate": 2e-05,
      "loss": 0.553,
      "step": 760
    },
    {
      "epoch": 0.06767445948321321,
      "grad_norm": 0.3414924740791321,
      "learning_rate": 2e-05,
      "loss": 0.5935,
      "step": 770
    },
    {
      "epoch": 0.0685533485674108,
      "grad_norm": 0.32405906915664673,
      "learning_rate": 2e-05,
      "loss": 0.5272,
      "step": 780
    },
    {
      "epoch": 0.06943223765160837,
      "grad_norm": 0.3730144500732422,
      "learning_rate": 2e-05,
      "loss": 0.5562,
      "step": 790
    },
    {
      "epoch": 0.07031112673580595,
      "grad_norm": 0.4821453392505646,
      "learning_rate": 2e-05,
      "loss": 0.575,
      "step": 800
    },
    {
      "epoch": 0.07119001582000352,
      "grad_norm": 0.48737847805023193,
      "learning_rate": 2e-05,
      "loss": 0.6033,
      "step": 810
    },
    {
      "epoch": 0.07206890490420109,
      "grad_norm": 0.5840171575546265,
      "learning_rate": 2e-05,
      "loss": 0.5745,
      "step": 820
    },
    {
      "epoch": 0.07294779398839867,
      "grad_norm": 0.3971763551235199,
      "learning_rate": 2e-05,
      "loss": 0.5696,
      "step": 830
    },
    {
      "epoch": 0.07382668307259624,
      "grad_norm": 0.3556879460811615,
      "learning_rate": 2e-05,
      "loss": 0.5993,
      "step": 840
    },
    {
      "epoch": 0.07470557215679381,
      "grad_norm": 0.3166077733039856,
      "learning_rate": 2e-05,
      "loss": 0.5711,
      "step": 850
    },
    {
      "epoch": 0.07558446124099139,
      "grad_norm": 0.37802746891975403,
      "learning_rate": 2e-05,
      "loss": 0.5806,
      "step": 860
    },
    {
      "epoch": 0.07646335032518896,
      "grad_norm": 0.33548033237457275,
      "learning_rate": 2e-05,
      "loss": 0.5619,
      "step": 870
    },
    {
      "epoch": 0.07734223940938653,
      "grad_norm": 0.3750935196876526,
      "learning_rate": 2e-05,
      "loss": 0.5636,
      "step": 880
    },
    {
      "epoch": 0.0782211284935841,
      "grad_norm": 0.359465628862381,
      "learning_rate": 2e-05,
      "loss": 0.5978,
      "step": 890
    },
    {
      "epoch": 0.07910001757778168,
      "grad_norm": 0.39016494154930115,
      "learning_rate": 2e-05,
      "loss": 0.5784,
      "step": 900
    },
    {
      "epoch": 0.07997890666197925,
      "grad_norm": 0.32137224078178406,
      "learning_rate": 2e-05,
      "loss": 0.5958,
      "step": 910
    },
    {
      "epoch": 0.08085779574617684,
      "grad_norm": 0.3473421037197113,
      "learning_rate": 2e-05,
      "loss": 0.5707,
      "step": 920
    },
    {
      "epoch": 0.08173668483037441,
      "grad_norm": 0.35025525093078613,
      "learning_rate": 2e-05,
      "loss": 0.5776,
      "step": 930
    },
    {
      "epoch": 0.08261557391457199,
      "grad_norm": 0.4126445949077606,
      "learning_rate": 2e-05,
      "loss": 0.5645,
      "step": 940
    },
    {
      "epoch": 0.08349446299876956,
      "grad_norm": 0.3999921679496765,
      "learning_rate": 2e-05,
      "loss": 0.5806,
      "step": 950
    },
    {
      "epoch": 0.08437335208296713,
      "grad_norm": 0.38202181458473206,
      "learning_rate": 2e-05,
      "loss": 0.5775,
      "step": 960
    },
    {
      "epoch": 0.0852522411671647,
      "grad_norm": 0.3595036566257477,
      "learning_rate": 2e-05,
      "loss": 0.5797,
      "step": 970
    },
    {
      "epoch": 0.08613113025136228,
      "grad_norm": 0.3591906726360321,
      "learning_rate": 2e-05,
      "loss": 0.5875,
      "step": 980
    },
    {
      "epoch": 0.08701001933555985,
      "grad_norm": 0.3604753613471985,
      "learning_rate": 2e-05,
      "loss": 0.576,
      "step": 990
    },
    {
      "epoch": 0.08788890841975742,
      "grad_norm": 0.35856151580810547,
      "learning_rate": 2e-05,
      "loss": 0.5704,
      "step": 1000
    },
    {
      "epoch": 0.088767797503955,
      "grad_norm": 0.3498944938182831,
      "learning_rate": 2e-05,
      "loss": 0.6101,
      "step": 1010
    },
    {
      "epoch": 0.08964668658815257,
      "grad_norm": 0.34088367223739624,
      "learning_rate": 2e-05,
      "loss": 0.5809,
      "step": 1020
    },
    {
      "epoch": 0.09052557567235014,
      "grad_norm": 0.37535375356674194,
      "learning_rate": 2e-05,
      "loss": 0.569,
      "step": 1030
    },
    {
      "epoch": 0.09140446475654772,
      "grad_norm": 0.34113436937332153,
      "learning_rate": 2e-05,
      "loss": 0.5821,
      "step": 1040
    },
    {
      "epoch": 0.09228335384074529,
      "grad_norm": 0.3853510618209839,
      "learning_rate": 2e-05,
      "loss": 0.6217,
      "step": 1050
    },
    {
      "epoch": 0.09316224292494288,
      "grad_norm": 0.39107996225357056,
      "learning_rate": 2e-05,
      "loss": 0.5619,
      "step": 1060
    },
    {
      "epoch": 0.09404113200914045,
      "grad_norm": 0.3326006531715393,
      "learning_rate": 2e-05,
      "loss": 0.5712,
      "step": 1070
    },
    {
      "epoch": 0.09492002109333803,
      "grad_norm": 0.5044306516647339,
      "learning_rate": 2e-05,
      "loss": 0.5799,
      "step": 1080
    },
    {
      "epoch": 0.0957989101775356,
      "grad_norm": 0.4056992530822754,
      "learning_rate": 2e-05,
      "loss": 0.5592,
      "step": 1090
    },
    {
      "epoch": 0.09667779926173317,
      "grad_norm": 0.42873504757881165,
      "learning_rate": 2e-05,
      "loss": 0.5932,
      "step": 1100
    },
    {
      "epoch": 0.09755668834593074,
      "grad_norm": 0.3402392566204071,
      "learning_rate": 2e-05,
      "loss": 0.5675,
      "step": 1110
    },
    {
      "epoch": 0.09843557743012832,
      "grad_norm": 0.36879539489746094,
      "learning_rate": 2e-05,
      "loss": 0.5393,
      "step": 1120
    },
    {
      "epoch": 0.09931446651432589,
      "grad_norm": 0.3497093617916107,
      "learning_rate": 2e-05,
      "loss": 0.6054,
      "step": 1130
    },
    {
      "epoch": 0.10019335559852346,
      "grad_norm": 0.3965706527233124,
      "learning_rate": 2e-05,
      "loss": 0.5851,
      "step": 1140
    },
    {
      "epoch": 0.10107224468272104,
      "grad_norm": 0.35941794514656067,
      "learning_rate": 2e-05,
      "loss": 0.5703,
      "step": 1150
    },
    {
      "epoch": 0.10195113376691861,
      "grad_norm": 0.3452138900756836,
      "learning_rate": 2e-05,
      "loss": 0.5854,
      "step": 1160
    },
    {
      "epoch": 0.10283002285111618,
      "grad_norm": 0.38294243812561035,
      "learning_rate": 2e-05,
      "loss": 0.5553,
      "step": 1170
    },
    {
      "epoch": 0.10370891193531376,
      "grad_norm": 0.3560450077056885,
      "learning_rate": 2e-05,
      "loss": 0.5847,
      "step": 1180
    },
    {
      "epoch": 0.10458780101951133,
      "grad_norm": 0.34364673495292664,
      "learning_rate": 2e-05,
      "loss": 0.5627,
      "step": 1190
    },
    {
      "epoch": 0.10546669010370892,
      "grad_norm": 0.3834216892719269,
      "learning_rate": 2e-05,
      "loss": 0.5617,
      "step": 1200
    },
    {
      "epoch": 0.10634557918790649,
      "grad_norm": 0.34551456570625305,
      "learning_rate": 2e-05,
      "loss": 0.555,
      "step": 1210
    },
    {
      "epoch": 0.10722446827210406,
      "grad_norm": 0.3550628423690796,
      "learning_rate": 2e-05,
      "loss": 0.569,
      "step": 1220
    },
    {
      "epoch": 0.10810335735630164,
      "grad_norm": 0.39383211731910706,
      "learning_rate": 2e-05,
      "loss": 0.5475,
      "step": 1230
    },
    {
      "epoch": 0.10898224644049921,
      "grad_norm": 0.45007845759391785,
      "learning_rate": 2e-05,
      "loss": 0.5749,
      "step": 1240
    },
    {
      "epoch": 0.10986113552469678,
      "grad_norm": 0.33808740973472595,
      "learning_rate": 2e-05,
      "loss": 0.5613,
      "step": 1250
    },
    {
      "epoch": 0.11074002460889436,
      "grad_norm": 0.3664572238922119,
      "learning_rate": 2e-05,
      "loss": 0.5525,
      "step": 1260
    },
    {
      "epoch": 0.11161891369309193,
      "grad_norm": 0.3647216260433197,
      "learning_rate": 2e-05,
      "loss": 0.5793,
      "step": 1270
    },
    {
      "epoch": 0.1124978027772895,
      "grad_norm": 0.3669637441635132,
      "learning_rate": 2e-05,
      "loss": 0.5572,
      "step": 1280
    },
    {
      "epoch": 0.11337669186148708,
      "grad_norm": 0.3178112506866455,
      "learning_rate": 2e-05,
      "loss": 0.5509,
      "step": 1290
    },
    {
      "epoch": 0.11425558094568465,
      "grad_norm": 0.360220342874527,
      "learning_rate": 2e-05,
      "loss": 0.5589,
      "step": 1300
    },
    {
      "epoch": 0.11513447002988222,
      "grad_norm": 0.36686965823173523,
      "learning_rate": 2e-05,
      "loss": 0.5752,
      "step": 1310
    },
    {
      "epoch": 0.1160133591140798,
      "grad_norm": 0.357736200094223,
      "learning_rate": 2e-05,
      "loss": 0.5901,
      "step": 1320
    },
    {
      "epoch": 0.11689224819827737,
      "grad_norm": 0.3455224633216858,
      "learning_rate": 2e-05,
      "loss": 0.5614,
      "step": 1330
    },
    {
      "epoch": 0.11777113728247496,
      "grad_norm": 0.34142857789993286,
      "learning_rate": 2e-05,
      "loss": 0.5563,
      "step": 1340
    },
    {
      "epoch": 0.11865002636667253,
      "grad_norm": 0.33547478914260864,
      "learning_rate": 2e-05,
      "loss": 0.5347,
      "step": 1350
    },
    {
      "epoch": 0.1195289154508701,
      "grad_norm": 0.3595579266548157,
      "learning_rate": 2e-05,
      "loss": 0.5747,
      "step": 1360
    },
    {
      "epoch": 0.12040780453506768,
      "grad_norm": 0.4105367362499237,
      "learning_rate": 2e-05,
      "loss": 0.5848,
      "step": 1370
    },
    {
      "epoch": 0.12128669361926525,
      "grad_norm": 0.45205193758010864,
      "learning_rate": 2e-05,
      "loss": 0.5863,
      "step": 1380
    },
    {
      "epoch": 0.12216558270346282,
      "grad_norm": 0.3702628016471863,
      "learning_rate": 2e-05,
      "loss": 0.5727,
      "step": 1390
    },
    {
      "epoch": 0.1230444717876604,
      "grad_norm": 0.3627513647079468,
      "learning_rate": 2e-05,
      "loss": 0.554,
      "step": 1400
    },
    {
      "epoch": 0.12392336087185797,
      "grad_norm": 0.37213680148124695,
      "learning_rate": 2e-05,
      "loss": 0.5441,
      "step": 1410
    },
    {
      "epoch": 0.12480224995605554,
      "grad_norm": 0.33166980743408203,
      "learning_rate": 2e-05,
      "loss": 0.5596,
      "step": 1420
    },
    {
      "epoch": 0.12568113904025313,
      "grad_norm": 0.36080440878868103,
      "learning_rate": 2e-05,
      "loss": 0.605,
      "step": 1430
    },
    {
      "epoch": 0.1265600281244507,
      "grad_norm": 0.39948782324790955,
      "learning_rate": 2e-05,
      "loss": 0.5642,
      "step": 1440
    },
    {
      "epoch": 0.12743891720864828,
      "grad_norm": 0.3613622784614563,
      "learning_rate": 2e-05,
      "loss": 0.5624,
      "step": 1450
    },
    {
      "epoch": 0.12831780629284584,
      "grad_norm": 0.41446202993392944,
      "learning_rate": 2e-05,
      "loss": 0.5966,
      "step": 1460
    },
    {
      "epoch": 0.12919669537704342,
      "grad_norm": 0.3632044196128845,
      "learning_rate": 2e-05,
      "loss": 0.558,
      "step": 1470
    },
    {
      "epoch": 0.13007558446124098,
      "grad_norm": 0.3540528118610382,
      "learning_rate": 2e-05,
      "loss": 0.5239,
      "step": 1480
    },
    {
      "epoch": 0.13095447354543857,
      "grad_norm": 0.43492934107780457,
      "learning_rate": 2e-05,
      "loss": 0.5639,
      "step": 1490
    },
    {
      "epoch": 0.13183336262963613,
      "grad_norm": 0.3482060432434082,
      "learning_rate": 2e-05,
      "loss": 0.5946,
      "step": 1500
    },
    {
      "epoch": 0.13271225171383372,
      "grad_norm": 0.34748104214668274,
      "learning_rate": 2e-05,
      "loss": 0.5628,
      "step": 1510
    },
    {
      "epoch": 0.13359114079803128,
      "grad_norm": 0.3222218453884125,
      "learning_rate": 2e-05,
      "loss": 0.5933,
      "step": 1520
    },
    {
      "epoch": 0.13447002988222886,
      "grad_norm": 0.3792349398136139,
      "learning_rate": 2e-05,
      "loss": 0.562,
      "step": 1530
    },
    {
      "epoch": 0.13534891896642642,
      "grad_norm": 0.37555038928985596,
      "learning_rate": 2e-05,
      "loss": 0.567,
      "step": 1540
    },
    {
      "epoch": 0.136227808050624,
      "grad_norm": 0.3451753854751587,
      "learning_rate": 2e-05,
      "loss": 0.5585,
      "step": 1550
    },
    {
      "epoch": 0.1371066971348216,
      "grad_norm": 0.32497790455818176,
      "learning_rate": 2e-05,
      "loss": 0.5659,
      "step": 1560
    },
    {
      "epoch": 0.13798558621901916,
      "grad_norm": 0.35301199555397034,
      "learning_rate": 2e-05,
      "loss": 0.6012,
      "step": 1570
    },
    {
      "epoch": 0.13886447530321674,
      "grad_norm": 0.3340965807437897,
      "learning_rate": 2e-05,
      "loss": 0.6125,
      "step": 1580
    },
    {
      "epoch": 0.1397433643874143,
      "grad_norm": 0.3181551396846771,
      "learning_rate": 2e-05,
      "loss": 0.5529,
      "step": 1590
    },
    {
      "epoch": 0.1406222534716119,
      "grad_norm": 0.3974955677986145,
      "learning_rate": 2e-05,
      "loss": 0.585,
      "step": 1600
    },
    {
      "epoch": 0.14150114255580945,
      "grad_norm": 0.347336083650589,
      "learning_rate": 2e-05,
      "loss": 0.5606,
      "step": 1610
    },
    {
      "epoch": 0.14238003164000704,
      "grad_norm": 0.3495638072490692,
      "learning_rate": 2e-05,
      "loss": 0.5906,
      "step": 1620
    },
    {
      "epoch": 0.1432589207242046,
      "grad_norm": 0.3735026717185974,
      "learning_rate": 2e-05,
      "loss": 0.5591,
      "step": 1630
    },
    {
      "epoch": 0.14413780980840218,
      "grad_norm": 0.3471795618534088,
      "learning_rate": 2e-05,
      "loss": 0.5705,
      "step": 1640
    },
    {
      "epoch": 0.14501669889259974,
      "grad_norm": 0.36941561102867126,
      "learning_rate": 2e-05,
      "loss": 0.6011,
      "step": 1650
    },
    {
      "epoch": 0.14589558797679733,
      "grad_norm": 0.3396647870540619,
      "learning_rate": 2e-05,
      "loss": 0.5717,
      "step": 1660
    },
    {
      "epoch": 0.1467744770609949,
      "grad_norm": 0.31963610649108887,
      "learning_rate": 2e-05,
      "loss": 0.5655,
      "step": 1670
    },
    {
      "epoch": 0.14765336614519248,
      "grad_norm": 0.44198107719421387,
      "learning_rate": 2e-05,
      "loss": 0.5869,
      "step": 1680
    },
    {
      "epoch": 0.14853225522939006,
      "grad_norm": 0.33693188428878784,
      "learning_rate": 2e-05,
      "loss": 0.5459,
      "step": 1690
    },
    {
      "epoch": 0.14941114431358762,
      "grad_norm": 0.3307777941226959,
      "learning_rate": 2e-05,
      "loss": 0.5901,
      "step": 1700
    },
    {
      "epoch": 0.1502900333977852,
      "grad_norm": 0.3152621388435364,
      "learning_rate": 2e-05,
      "loss": 0.5596,
      "step": 1710
    },
    {
      "epoch": 0.15116892248198277,
      "grad_norm": 0.5023934841156006,
      "learning_rate": 2e-05,
      "loss": 0.5634,
      "step": 1720
    },
    {
      "epoch": 0.15204781156618036,
      "grad_norm": 0.3895277678966522,
      "learning_rate": 2e-05,
      "loss": 0.5763,
      "step": 1730
    },
    {
      "epoch": 0.15292670065037792,
      "grad_norm": 0.3810598850250244,
      "learning_rate": 2e-05,
      "loss": 0.576,
      "step": 1740
    },
    {
      "epoch": 0.1538055897345755,
      "grad_norm": 0.3689657151699066,
      "learning_rate": 2e-05,
      "loss": 0.5474,
      "step": 1750
    },
    {
      "epoch": 0.15468447881877306,
      "grad_norm": 0.3715059459209442,
      "learning_rate": 2e-05,
      "loss": 0.5491,
      "step": 1760
    },
    {
      "epoch": 0.15556336790297065,
      "grad_norm": 0.33421632647514343,
      "learning_rate": 2e-05,
      "loss": 0.5425,
      "step": 1770
    },
    {
      "epoch": 0.1564422569871682,
      "grad_norm": 0.5304133296012878,
      "learning_rate": 2e-05,
      "loss": 0.5716,
      "step": 1780
    },
    {
      "epoch": 0.1573211460713658,
      "grad_norm": 0.3863163888454437,
      "learning_rate": 2e-05,
      "loss": 0.5795,
      "step": 1790
    },
    {
      "epoch": 0.15820003515556336,
      "grad_norm": 0.5527595281600952,
      "learning_rate": 2e-05,
      "loss": 0.5561,
      "step": 1800
    },
    {
      "epoch": 0.15907892423976094,
      "grad_norm": 0.4207919239997864,
      "learning_rate": 2e-05,
      "loss": 0.5837,
      "step": 1810
    },
    {
      "epoch": 0.1599578133239585,
      "grad_norm": 0.29921954870224,
      "learning_rate": 2e-05,
      "loss": 0.5299,
      "step": 1820
    },
    {
      "epoch": 0.1608367024081561,
      "grad_norm": 0.34096500277519226,
      "learning_rate": 2e-05,
      "loss": 0.5767,
      "step": 1830
    },
    {
      "epoch": 0.16171559149235368,
      "grad_norm": 0.37093764543533325,
      "learning_rate": 2e-05,
      "loss": 0.5597,
      "step": 1840
    },
    {
      "epoch": 0.16259448057655124,
      "grad_norm": 0.392698734998703,
      "learning_rate": 2e-05,
      "loss": 0.5546,
      "step": 1850
    },
    {
      "epoch": 0.16347336966074882,
      "grad_norm": 0.37084656953811646,
      "learning_rate": 2e-05,
      "loss": 0.564,
      "step": 1860
    },
    {
      "epoch": 0.16435225874494638,
      "grad_norm": 0.5617843866348267,
      "learning_rate": 2e-05,
      "loss": 0.5669,
      "step": 1870
    },
    {
      "epoch": 0.16523114782914397,
      "grad_norm": 0.3692229390144348,
      "learning_rate": 2e-05,
      "loss": 0.5628,
      "step": 1880
    },
    {
      "epoch": 0.16611003691334153,
      "grad_norm": 0.3360963761806488,
      "learning_rate": 2e-05,
      "loss": 0.551,
      "step": 1890
    },
    {
      "epoch": 0.16698892599753912,
      "grad_norm": 0.3678745925426483,
      "learning_rate": 2e-05,
      "loss": 0.5376,
      "step": 1900
    },
    {
      "epoch": 0.16786781508173668,
      "grad_norm": 0.3097802996635437,
      "learning_rate": 2e-05,
      "loss": 0.5924,
      "step": 1910
    },
    {
      "epoch": 0.16874670416593426,
      "grad_norm": 0.3742968738079071,
      "learning_rate": 2e-05,
      "loss": 0.5538,
      "step": 1920
    },
    {
      "epoch": 0.16962559325013182,
      "grad_norm": 0.3674496114253998,
      "learning_rate": 2e-05,
      "loss": 0.5837,
      "step": 1930
    },
    {
      "epoch": 0.1705044823343294,
      "grad_norm": 0.41291990876197815,
      "learning_rate": 2e-05,
      "loss": 0.5453,
      "step": 1940
    },
    {
      "epoch": 0.17138337141852697,
      "grad_norm": 0.35975903272628784,
      "learning_rate": 2e-05,
      "loss": 0.5478,
      "step": 1950
    },
    {
      "epoch": 0.17226226050272456,
      "grad_norm": 0.31317245960235596,
      "learning_rate": 2e-05,
      "loss": 0.5891,
      "step": 1960
    },
    {
      "epoch": 0.17314114958692214,
      "grad_norm": 0.3947499394416809,
      "learning_rate": 2e-05,
      "loss": 0.5555,
      "step": 1970
    },
    {
      "epoch": 0.1740200386711197,
      "grad_norm": 0.41339772939682007,
      "learning_rate": 2e-05,
      "loss": 0.5683,
      "step": 1980
    },
    {
      "epoch": 0.1748989277553173,
      "grad_norm": 0.3648429811000824,
      "learning_rate": 2e-05,
      "loss": 0.5639,
      "step": 1990
    },
    {
      "epoch": 0.17577781683951485,
      "grad_norm": 0.31032198667526245,
      "learning_rate": 2e-05,
      "loss": 0.5613,
      "step": 2000
    },
    {
      "epoch": 0.17665670592371244,
      "grad_norm": 0.3212377429008484,
      "learning_rate": 2e-05,
      "loss": 0.5501,
      "step": 2010
    },
    {
      "epoch": 0.17753559500791,
      "grad_norm": 0.34276482462882996,
      "learning_rate": 2e-05,
      "loss": 0.5801,
      "step": 2020
    },
    {
      "epoch": 0.17841448409210758,
      "grad_norm": 0.32408374547958374,
      "learning_rate": 2e-05,
      "loss": 0.574,
      "step": 2030
    },
    {
      "epoch": 0.17929337317630514,
      "grad_norm": 0.31184566020965576,
      "learning_rate": 2e-05,
      "loss": 0.5566,
      "step": 2040
    },
    {
      "epoch": 0.18017226226050273,
      "grad_norm": 0.32635489106178284,
      "learning_rate": 2e-05,
      "loss": 0.5475,
      "step": 2050
    },
    {
      "epoch": 0.1810511513447003,
      "grad_norm": 0.3519839644432068,
      "learning_rate": 2e-05,
      "loss": 0.5652,
      "step": 2060
    },
    {
      "epoch": 0.18193004042889788,
      "grad_norm": 0.37629371881484985,
      "learning_rate": 2e-05,
      "loss": 0.5521,
      "step": 2070
    },
    {
      "epoch": 0.18280892951309544,
      "grad_norm": 0.3756372034549713,
      "learning_rate": 2e-05,
      "loss": 0.5621,
      "step": 2080
    },
    {
      "epoch": 0.18368781859729302,
      "grad_norm": 0.3752441704273224,
      "learning_rate": 2e-05,
      "loss": 0.5802,
      "step": 2090
    },
    {
      "epoch": 0.18456670768149058,
      "grad_norm": 0.3569568693637848,
      "learning_rate": 2e-05,
      "loss": 0.5828,
      "step": 2100
    },
    {
      "epoch": 0.18544559676568817,
      "grad_norm": 0.32334834337234497,
      "learning_rate": 2e-05,
      "loss": 0.5479,
      "step": 2110
    },
    {
      "epoch": 0.18632448584988576,
      "grad_norm": 0.35802629590034485,
      "learning_rate": 2e-05,
      "loss": 0.566,
      "step": 2120
    },
    {
      "epoch": 0.18720337493408332,
      "grad_norm": 0.5078518986701965,
      "learning_rate": 2e-05,
      "loss": 0.5586,
      "step": 2130
    },
    {
      "epoch": 0.1880822640182809,
      "grad_norm": 0.36526328325271606,
      "learning_rate": 2e-05,
      "loss": 0.5616,
      "step": 2140
    },
    {
      "epoch": 0.18896115310247846,
      "grad_norm": 0.33778101205825806,
      "learning_rate": 2e-05,
      "loss": 0.5666,
      "step": 2150
    },
    {
      "epoch": 0.18984004218667605,
      "grad_norm": 0.38717159628868103,
      "learning_rate": 2e-05,
      "loss": 0.5605,
      "step": 2160
    },
    {
      "epoch": 0.1907189312708736,
      "grad_norm": 0.325623095035553,
      "learning_rate": 2e-05,
      "loss": 0.5419,
      "step": 2170
    },
    {
      "epoch": 0.1915978203550712,
      "grad_norm": 0.4406554400920868,
      "learning_rate": 2e-05,
      "loss": 0.5856,
      "step": 2180
    },
    {
      "epoch": 0.19247670943926876,
      "grad_norm": 0.3608305752277374,
      "learning_rate": 2e-05,
      "loss": 0.5397,
      "step": 2190
    },
    {
      "epoch": 0.19335559852346634,
      "grad_norm": 0.4066288471221924,
      "learning_rate": 2e-05,
      "loss": 0.5884,
      "step": 2200
    },
    {
      "epoch": 0.1942344876076639,
      "grad_norm": 0.37264975905418396,
      "learning_rate": 2e-05,
      "loss": 0.6049,
      "step": 2210
    },
    {
      "epoch": 0.1951133766918615,
      "grad_norm": 0.3412500321865082,
      "learning_rate": 2e-05,
      "loss": 0.5679,
      "step": 2220
    },
    {
      "epoch": 0.19599226577605905,
      "grad_norm": 0.36977285146713257,
      "learning_rate": 2e-05,
      "loss": 0.5589,
      "step": 2230
    },
    {
      "epoch": 0.19687115486025664,
      "grad_norm": 0.320394903421402,
      "learning_rate": 2e-05,
      "loss": 0.5536,
      "step": 2240
    },
    {
      "epoch": 0.19775004394445422,
      "grad_norm": 0.4130312502384186,
      "learning_rate": 2e-05,
      "loss": 0.5678,
      "step": 2250
    },
    {
      "epoch": 0.19862893302865178,
      "grad_norm": 0.3542708158493042,
      "learning_rate": 2e-05,
      "loss": 0.5871,
      "step": 2260
    },
    {
      "epoch": 0.19950782211284937,
      "grad_norm": 0.36599817872047424,
      "learning_rate": 2e-05,
      "loss": 0.5481,
      "step": 2270
    },
    {
      "epoch": 0.20038671119704693,
      "grad_norm": 0.42361220717430115,
      "learning_rate": 2e-05,
      "loss": 0.546,
      "step": 2280
    },
    {
      "epoch": 0.20126560028124452,
      "grad_norm": 0.3402586579322815,
      "learning_rate": 2e-05,
      "loss": 0.58,
      "step": 2290
    },
    {
      "epoch": 0.20214448936544208,
      "grad_norm": 0.3884638845920563,
      "learning_rate": 2e-05,
      "loss": 0.5393,
      "step": 2300
    },
    {
      "epoch": 0.20302337844963966,
      "grad_norm": 0.366931676864624,
      "learning_rate": 2e-05,
      "loss": 0.596,
      "step": 2310
    },
    {
      "epoch": 0.20390226753383722,
      "grad_norm": 0.32991132140159607,
      "learning_rate": 2e-05,
      "loss": 0.5522,
      "step": 2320
    },
    {
      "epoch": 0.2047811566180348,
      "grad_norm": 0.39930519461631775,
      "learning_rate": 2e-05,
      "loss": 0.5836,
      "step": 2330
    },
    {
      "epoch": 0.20566004570223237,
      "grad_norm": 0.37963399291038513,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 2340
    },
    {
      "epoch": 0.20653893478642996,
      "grad_norm": 0.365708589553833,
      "learning_rate": 2e-05,
      "loss": 0.5646,
      "step": 2350
    },
    {
      "epoch": 0.20741782387062752,
      "grad_norm": 0.38639870285987854,
      "learning_rate": 2e-05,
      "loss": 0.5348,
      "step": 2360
    },
    {
      "epoch": 0.2082967129548251,
      "grad_norm": 0.31267666816711426,
      "learning_rate": 2e-05,
      "loss": 0.541,
      "step": 2370
    },
    {
      "epoch": 0.20917560203902266,
      "grad_norm": 0.42928454279899597,
      "learning_rate": 2e-05,
      "loss": 0.5665,
      "step": 2380
    },
    {
      "epoch": 0.21005449112322025,
      "grad_norm": 0.3244638442993164,
      "learning_rate": 2e-05,
      "loss": 0.5904,
      "step": 2390
    },
    {
      "epoch": 0.21093338020741784,
      "grad_norm": 0.30141741037368774,
      "learning_rate": 2e-05,
      "loss": 0.5446,
      "step": 2400
    },
    {
      "epoch": 0.2118122692916154,
      "grad_norm": 0.3784339427947998,
      "learning_rate": 2e-05,
      "loss": 0.5936,
      "step": 2410
    },
    {
      "epoch": 0.21269115837581298,
      "grad_norm": 0.40315935015678406,
      "learning_rate": 2e-05,
      "loss": 0.5707,
      "step": 2420
    },
    {
      "epoch": 0.21357004746001054,
      "grad_norm": 0.35042673349380493,
      "learning_rate": 2e-05,
      "loss": 0.5684,
      "step": 2430
    },
    {
      "epoch": 0.21444893654420813,
      "grad_norm": 0.37262678146362305,
      "learning_rate": 2e-05,
      "loss": 0.5531,
      "step": 2440
    },
    {
      "epoch": 0.2153278256284057,
      "grad_norm": 0.3180401027202606,
      "learning_rate": 2e-05,
      "loss": 0.5727,
      "step": 2450
    },
    {
      "epoch": 0.21620671471260328,
      "grad_norm": 0.3344835937023163,
      "learning_rate": 2e-05,
      "loss": 0.5707,
      "step": 2460
    },
    {
      "epoch": 0.21708560379680084,
      "grad_norm": 0.3284481167793274,
      "learning_rate": 2e-05,
      "loss": 0.5497,
      "step": 2470
    },
    {
      "epoch": 0.21796449288099842,
      "grad_norm": 0.37181684374809265,
      "learning_rate": 2e-05,
      "loss": 0.5904,
      "step": 2480
    },
    {
      "epoch": 0.21884338196519598,
      "grad_norm": 0.3416779637336731,
      "learning_rate": 2e-05,
      "loss": 0.5507,
      "step": 2490
    },
    {
      "epoch": 0.21972227104939357,
      "grad_norm": 0.347848504781723,
      "learning_rate": 2e-05,
      "loss": 0.5684,
      "step": 2500
    },
    {
      "epoch": 0.22060116013359113,
      "grad_norm": 0.3230196237564087,
      "learning_rate": 2e-05,
      "loss": 0.5456,
      "step": 2510
    },
    {
      "epoch": 0.22148004921778872,
      "grad_norm": 0.38809800148010254,
      "learning_rate": 2e-05,
      "loss": 0.5517,
      "step": 2520
    },
    {
      "epoch": 0.2223589383019863,
      "grad_norm": 0.3716602027416229,
      "learning_rate": 2e-05,
      "loss": 0.5606,
      "step": 2530
    },
    {
      "epoch": 0.22323782738618386,
      "grad_norm": 0.342180460691452,
      "learning_rate": 2e-05,
      "loss": 0.55,
      "step": 2540
    },
    {
      "epoch": 0.22411671647038145,
      "grad_norm": 0.40020951628685,
      "learning_rate": 2e-05,
      "loss": 0.5326,
      "step": 2550
    },
    {
      "epoch": 0.224995605554579,
      "grad_norm": 0.37780147790908813,
      "learning_rate": 2e-05,
      "loss": 0.5473,
      "step": 2560
    },
    {
      "epoch": 0.2258744946387766,
      "grad_norm": 0.3578275740146637,
      "learning_rate": 2e-05,
      "loss": 0.5474,
      "step": 2570
    },
    {
      "epoch": 0.22675338372297416,
      "grad_norm": 0.33693841099739075,
      "learning_rate": 2e-05,
      "loss": 0.5686,
      "step": 2580
    },
    {
      "epoch": 0.22763227280717174,
      "grad_norm": 0.41994157433509827,
      "learning_rate": 2e-05,
      "loss": 0.5629,
      "step": 2590
    },
    {
      "epoch": 0.2285111618913693,
      "grad_norm": 0.3703048527240753,
      "learning_rate": 2e-05,
      "loss": 0.5535,
      "step": 2600
    },
    {
      "epoch": 0.2293900509755669,
      "grad_norm": 0.3215305805206299,
      "learning_rate": 2e-05,
      "loss": 0.5304,
      "step": 2610
    },
    {
      "epoch": 0.23026894005976445,
      "grad_norm": 0.3829116225242615,
      "learning_rate": 2e-05,
      "loss": 0.5717,
      "step": 2620
    },
    {
      "epoch": 0.23114782914396204,
      "grad_norm": 0.31442129611968994,
      "learning_rate": 2e-05,
      "loss": 0.5543,
      "step": 2630
    },
    {
      "epoch": 0.2320267182281596,
      "grad_norm": 0.46074509620666504,
      "learning_rate": 2e-05,
      "loss": 0.5811,
      "step": 2640
    },
    {
      "epoch": 0.23290560731235718,
      "grad_norm": 0.37541741132736206,
      "learning_rate": 2e-05,
      "loss": 0.5561,
      "step": 2650
    },
    {
      "epoch": 0.23378449639655474,
      "grad_norm": 0.32522743940353394,
      "learning_rate": 2e-05,
      "loss": 0.5649,
      "step": 2660
    },
    {
      "epoch": 0.23466338548075233,
      "grad_norm": 0.3570707142353058,
      "learning_rate": 2e-05,
      "loss": 0.5615,
      "step": 2670
    },
    {
      "epoch": 0.23554227456494992,
      "grad_norm": 0.33403363823890686,
      "learning_rate": 2e-05,
      "loss": 0.5733,
      "step": 2680
    },
    {
      "epoch": 0.23642116364914748,
      "grad_norm": 0.3274398744106293,
      "learning_rate": 2e-05,
      "loss": 0.5624,
      "step": 2690
    },
    {
      "epoch": 0.23730005273334506,
      "grad_norm": 0.35692623257637024,
      "learning_rate": 2e-05,
      "loss": 0.5733,
      "step": 2700
    },
    {
      "epoch": 0.23817894181754262,
      "grad_norm": 0.3499510586261749,
      "learning_rate": 2e-05,
      "loss": 0.581,
      "step": 2710
    },
    {
      "epoch": 0.2390578309017402,
      "grad_norm": 0.3738914430141449,
      "learning_rate": 2e-05,
      "loss": 0.5481,
      "step": 2720
    },
    {
      "epoch": 0.23993671998593777,
      "grad_norm": 0.4006647765636444,
      "learning_rate": 2e-05,
      "loss": 0.5337,
      "step": 2730
    },
    {
      "epoch": 0.24081560907013536,
      "grad_norm": 0.5295825600624084,
      "learning_rate": 2e-05,
      "loss": 0.561,
      "step": 2740
    },
    {
      "epoch": 0.24169449815433292,
      "grad_norm": 0.34328094124794006,
      "learning_rate": 2e-05,
      "loss": 0.5517,
      "step": 2750
    },
    {
      "epoch": 0.2425733872385305,
      "grad_norm": 0.36072251200675964,
      "learning_rate": 2e-05,
      "loss": 0.5486,
      "step": 2760
    },
    {
      "epoch": 0.24345227632272806,
      "grad_norm": 0.32866302132606506,
      "learning_rate": 2e-05,
      "loss": 0.5764,
      "step": 2770
    },
    {
      "epoch": 0.24433116540692565,
      "grad_norm": 0.3333951532840729,
      "learning_rate": 2e-05,
      "loss": 0.5775,
      "step": 2780
    },
    {
      "epoch": 0.2452100544911232,
      "grad_norm": 0.41300830245018005,
      "learning_rate": 2e-05,
      "loss": 0.5609,
      "step": 2790
    },
    {
      "epoch": 0.2460889435753208,
      "grad_norm": 0.3127140700817108,
      "learning_rate": 2e-05,
      "loss": 0.5711,
      "step": 2800
    },
    {
      "epoch": 0.24696783265951838,
      "grad_norm": 0.33722373843193054,
      "learning_rate": 2e-05,
      "loss": 0.5421,
      "step": 2810
    },
    {
      "epoch": 0.24784672174371594,
      "grad_norm": 0.3545294404029846,
      "learning_rate": 2e-05,
      "loss": 0.568,
      "step": 2820
    },
    {
      "epoch": 0.24872561082791353,
      "grad_norm": 0.345797598361969,
      "learning_rate": 2e-05,
      "loss": 0.5306,
      "step": 2830
    },
    {
      "epoch": 0.2496044999121111,
      "grad_norm": 0.40061184763908386,
      "learning_rate": 2e-05,
      "loss": 0.5467,
      "step": 2840
    },
    {
      "epoch": 0.2504833889963087,
      "grad_norm": 0.32904642820358276,
      "learning_rate": 2e-05,
      "loss": 0.5294,
      "step": 2850
    },
    {
      "epoch": 0.25136227808050626,
      "grad_norm": 0.3608121871948242,
      "learning_rate": 2e-05,
      "loss": 0.5564,
      "step": 2860
    },
    {
      "epoch": 0.2522411671647038,
      "grad_norm": 0.4847877025604248,
      "learning_rate": 2e-05,
      "loss": 0.5627,
      "step": 2870
    },
    {
      "epoch": 0.2531200562489014,
      "grad_norm": 0.3074108362197876,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 2880
    },
    {
      "epoch": 0.25399894533309897,
      "grad_norm": 0.3400770127773285,
      "learning_rate": 2e-05,
      "loss": 0.5395,
      "step": 2890
    },
    {
      "epoch": 0.25487783441729656,
      "grad_norm": 0.3570431172847748,
      "learning_rate": 2e-05,
      "loss": 0.5627,
      "step": 2900
    },
    {
      "epoch": 0.2557567235014941,
      "grad_norm": 0.3542000651359558,
      "learning_rate": 2e-05,
      "loss": 0.5667,
      "step": 2910
    },
    {
      "epoch": 0.2566356125856917,
      "grad_norm": 0.3159409761428833,
      "learning_rate": 2e-05,
      "loss": 0.5561,
      "step": 2920
    },
    {
      "epoch": 0.25751450166988926,
      "grad_norm": 0.3355763554573059,
      "learning_rate": 2e-05,
      "loss": 0.553,
      "step": 2930
    },
    {
      "epoch": 0.25839339075408685,
      "grad_norm": 0.3804934024810791,
      "learning_rate": 2e-05,
      "loss": 0.5515,
      "step": 2940
    },
    {
      "epoch": 0.25927227983828444,
      "grad_norm": 0.3341805636882782,
      "learning_rate": 2e-05,
      "loss": 0.5695,
      "step": 2950
    },
    {
      "epoch": 0.26015116892248197,
      "grad_norm": 0.32842281460762024,
      "learning_rate": 2e-05,
      "loss": 0.5395,
      "step": 2960
    },
    {
      "epoch": 0.26103005800667956,
      "grad_norm": 0.3404848575592041,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 2970
    },
    {
      "epoch": 0.26190894709087714,
      "grad_norm": 0.38430777192115784,
      "learning_rate": 2e-05,
      "loss": 0.5637,
      "step": 2980
    },
    {
      "epoch": 0.26278783617507473,
      "grad_norm": 0.30766552686691284,
      "learning_rate": 2e-05,
      "loss": 0.5596,
      "step": 2990
    },
    {
      "epoch": 0.26366672525927226,
      "grad_norm": 0.351011723279953,
      "learning_rate": 2e-05,
      "loss": 0.5807,
      "step": 3000
    },
    {
      "epoch": 0.26454561434346985,
      "grad_norm": 0.35301610827445984,
      "learning_rate": 2e-05,
      "loss": 0.5583,
      "step": 3010
    },
    {
      "epoch": 0.26542450342766744,
      "grad_norm": 0.350064218044281,
      "learning_rate": 2e-05,
      "loss": 0.5608,
      "step": 3020
    },
    {
      "epoch": 0.266303392511865,
      "grad_norm": 0.3541603982448578,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 3030
    },
    {
      "epoch": 0.26718228159606255,
      "grad_norm": 0.3428955674171448,
      "learning_rate": 2e-05,
      "loss": 0.5658,
      "step": 3040
    },
    {
      "epoch": 0.26806117068026014,
      "grad_norm": 0.34655940532684326,
      "learning_rate": 2e-05,
      "loss": 0.5748,
      "step": 3050
    },
    {
      "epoch": 0.26894005976445773,
      "grad_norm": 0.3420259356498718,
      "learning_rate": 2e-05,
      "loss": 0.5447,
      "step": 3060
    },
    {
      "epoch": 0.2698189488486553,
      "grad_norm": 0.3394719958305359,
      "learning_rate": 2e-05,
      "loss": 0.5539,
      "step": 3070
    },
    {
      "epoch": 0.27069783793285285,
      "grad_norm": 0.33060818910598755,
      "learning_rate": 2e-05,
      "loss": 0.576,
      "step": 3080
    },
    {
      "epoch": 0.27157672701705043,
      "grad_norm": 0.34807607531547546,
      "learning_rate": 2e-05,
      "loss": 0.5524,
      "step": 3090
    },
    {
      "epoch": 0.272455616101248,
      "grad_norm": 0.33848682045936584,
      "learning_rate": 2e-05,
      "loss": 0.5283,
      "step": 3100
    },
    {
      "epoch": 0.2733345051854456,
      "grad_norm": 0.31702762842178345,
      "learning_rate": 2e-05,
      "loss": 0.5731,
      "step": 3110
    },
    {
      "epoch": 0.2742133942696432,
      "grad_norm": 0.30397921800613403,
      "learning_rate": 2e-05,
      "loss": 0.5403,
      "step": 3120
    },
    {
      "epoch": 0.2750922833538407,
      "grad_norm": 0.37567466497421265,
      "learning_rate": 2e-05,
      "loss": 0.5754,
      "step": 3130
    },
    {
      "epoch": 0.2759711724380383,
      "grad_norm": 0.3388182520866394,
      "learning_rate": 2e-05,
      "loss": 0.5512,
      "step": 3140
    },
    {
      "epoch": 0.2768500615222359,
      "grad_norm": 0.36109867691993713,
      "learning_rate": 2e-05,
      "loss": 0.5507,
      "step": 3150
    },
    {
      "epoch": 0.2777289506064335,
      "grad_norm": 0.3950864374637604,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 3160
    },
    {
      "epoch": 0.278607839690631,
      "grad_norm": 0.32862675189971924,
      "learning_rate": 2e-05,
      "loss": 0.5705,
      "step": 3170
    },
    {
      "epoch": 0.2794867287748286,
      "grad_norm": 0.3726232945919037,
      "learning_rate": 2e-05,
      "loss": 0.5456,
      "step": 3180
    },
    {
      "epoch": 0.2803656178590262,
      "grad_norm": 0.3371773958206177,
      "learning_rate": 2e-05,
      "loss": 0.5813,
      "step": 3190
    },
    {
      "epoch": 0.2812445069432238,
      "grad_norm": 0.3559187650680542,
      "learning_rate": 2e-05,
      "loss": 0.5193,
      "step": 3200
    },
    {
      "epoch": 0.2821233960274213,
      "grad_norm": 0.3417247235774994,
      "learning_rate": 2e-05,
      "loss": 0.5369,
      "step": 3210
    },
    {
      "epoch": 0.2830022851116189,
      "grad_norm": 0.3516131341457367,
      "learning_rate": 2e-05,
      "loss": 0.539,
      "step": 3220
    },
    {
      "epoch": 0.2838811741958165,
      "grad_norm": 0.29232025146484375,
      "learning_rate": 2e-05,
      "loss": 0.5746,
      "step": 3230
    },
    {
      "epoch": 0.2847600632800141,
      "grad_norm": 0.3936523199081421,
      "learning_rate": 2e-05,
      "loss": 0.5592,
      "step": 3240
    },
    {
      "epoch": 0.28563895236421166,
      "grad_norm": 0.3399224579334259,
      "learning_rate": 2e-05,
      "loss": 0.5602,
      "step": 3250
    },
    {
      "epoch": 0.2865178414484092,
      "grad_norm": 0.36498552560806274,
      "learning_rate": 2e-05,
      "loss": 0.5346,
      "step": 3260
    },
    {
      "epoch": 0.2873967305326068,
      "grad_norm": 0.3364720344543457,
      "learning_rate": 2e-05,
      "loss": 0.5615,
      "step": 3270
    },
    {
      "epoch": 0.28827561961680437,
      "grad_norm": 0.3698938190937042,
      "learning_rate": 2e-05,
      "loss": 0.5675,
      "step": 3280
    },
    {
      "epoch": 0.28915450870100196,
      "grad_norm": 0.3517701327800751,
      "learning_rate": 2e-05,
      "loss": 0.5494,
      "step": 3290
    },
    {
      "epoch": 0.2900333977851995,
      "grad_norm": 0.5226931571960449,
      "learning_rate": 2e-05,
      "loss": 0.5624,
      "step": 3300
    },
    {
      "epoch": 0.2909122868693971,
      "grad_norm": 0.36056846380233765,
      "learning_rate": 2e-05,
      "loss": 0.5549,
      "step": 3310
    },
    {
      "epoch": 0.29179117595359466,
      "grad_norm": 0.36480969190597534,
      "learning_rate": 2e-05,
      "loss": 0.555,
      "step": 3320
    },
    {
      "epoch": 0.29267006503779225,
      "grad_norm": 0.4041381776332855,
      "learning_rate": 2e-05,
      "loss": 0.5707,
      "step": 3330
    },
    {
      "epoch": 0.2935489541219898,
      "grad_norm": 0.3795798420906067,
      "learning_rate": 2e-05,
      "loss": 0.5224,
      "step": 3340
    },
    {
      "epoch": 0.29442784320618737,
      "grad_norm": 0.35352370142936707,
      "learning_rate": 2e-05,
      "loss": 0.5392,
      "step": 3350
    },
    {
      "epoch": 0.29530673229038495,
      "grad_norm": 0.4036523997783661,
      "learning_rate": 2e-05,
      "loss": 0.557,
      "step": 3360
    },
    {
      "epoch": 0.29618562137458254,
      "grad_norm": 0.3892285227775574,
      "learning_rate": 2e-05,
      "loss": 0.5704,
      "step": 3370
    },
    {
      "epoch": 0.29706451045878013,
      "grad_norm": 0.3575426936149597,
      "learning_rate": 2e-05,
      "loss": 0.5564,
      "step": 3380
    },
    {
      "epoch": 0.29794339954297766,
      "grad_norm": 0.35260751843452454,
      "learning_rate": 2e-05,
      "loss": 0.5423,
      "step": 3390
    },
    {
      "epoch": 0.29882228862717525,
      "grad_norm": 0.3310370445251465,
      "learning_rate": 2e-05,
      "loss": 0.5423,
      "step": 3400
    },
    {
      "epoch": 0.29970117771137283,
      "grad_norm": 0.35556313395500183,
      "learning_rate": 2e-05,
      "loss": 0.5498,
      "step": 3410
    },
    {
      "epoch": 0.3005800667955704,
      "grad_norm": 0.3494598865509033,
      "learning_rate": 2e-05,
      "loss": 0.532,
      "step": 3420
    },
    {
      "epoch": 0.30145895587976795,
      "grad_norm": 0.32750406861305237,
      "learning_rate": 2e-05,
      "loss": 0.5501,
      "step": 3430
    },
    {
      "epoch": 0.30233784496396554,
      "grad_norm": 0.34658825397491455,
      "learning_rate": 2e-05,
      "loss": 0.5513,
      "step": 3440
    },
    {
      "epoch": 0.30321673404816313,
      "grad_norm": 0.3852780759334564,
      "learning_rate": 2e-05,
      "loss": 0.5597,
      "step": 3450
    },
    {
      "epoch": 0.3040956231323607,
      "grad_norm": 0.3383525013923645,
      "learning_rate": 2e-05,
      "loss": 0.5396,
      "step": 3460
    },
    {
      "epoch": 0.30497451221655825,
      "grad_norm": 0.32666006684303284,
      "learning_rate": 2e-05,
      "loss": 0.5472,
      "step": 3470
    },
    {
      "epoch": 0.30585340130075583,
      "grad_norm": 0.3554231524467468,
      "learning_rate": 2e-05,
      "loss": 0.5587,
      "step": 3480
    },
    {
      "epoch": 0.3067322903849534,
      "grad_norm": 0.3399517834186554,
      "learning_rate": 2e-05,
      "loss": 0.5218,
      "step": 3490
    },
    {
      "epoch": 0.307611179469151,
      "grad_norm": 0.36986055970191956,
      "learning_rate": 2e-05,
      "loss": 0.5445,
      "step": 3500
    },
    {
      "epoch": 0.30849006855334854,
      "grad_norm": 0.3844973146915436,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 3510
    },
    {
      "epoch": 0.3093689576375461,
      "grad_norm": 0.3885951042175293,
      "learning_rate": 2e-05,
      "loss": 0.5599,
      "step": 3520
    },
    {
      "epoch": 0.3102478467217437,
      "grad_norm": 0.3187924325466156,
      "learning_rate": 2e-05,
      "loss": 0.5647,
      "step": 3530
    },
    {
      "epoch": 0.3111267358059413,
      "grad_norm": 0.3450549244880676,
      "learning_rate": 2e-05,
      "loss": 0.5657,
      "step": 3540
    },
    {
      "epoch": 0.3120056248901389,
      "grad_norm": 0.37729015946388245,
      "learning_rate": 2e-05,
      "loss": 0.5515,
      "step": 3550
    },
    {
      "epoch": 0.3128845139743364,
      "grad_norm": 0.3477257490158081,
      "learning_rate": 2e-05,
      "loss": 0.529,
      "step": 3560
    },
    {
      "epoch": 0.313763403058534,
      "grad_norm": 0.4619958698749542,
      "learning_rate": 2e-05,
      "loss": 0.5473,
      "step": 3570
    },
    {
      "epoch": 0.3146422921427316,
      "grad_norm": 0.4123147130012512,
      "learning_rate": 2e-05,
      "loss": 0.5699,
      "step": 3580
    },
    {
      "epoch": 0.3155211812269292,
      "grad_norm": 0.331781268119812,
      "learning_rate": 2e-05,
      "loss": 0.5477,
      "step": 3590
    },
    {
      "epoch": 0.3164000703111267,
      "grad_norm": 0.4143794775009155,
      "learning_rate": 2e-05,
      "loss": 0.5437,
      "step": 3600
    },
    {
      "epoch": 0.3172789593953243,
      "grad_norm": 0.37911897897720337,
      "learning_rate": 2e-05,
      "loss": 0.5698,
      "step": 3610
    },
    {
      "epoch": 0.3181578484795219,
      "grad_norm": 0.37786126136779785,
      "learning_rate": 2e-05,
      "loss": 0.5721,
      "step": 3620
    },
    {
      "epoch": 0.3190367375637195,
      "grad_norm": 0.36114272475242615,
      "learning_rate": 2e-05,
      "loss": 0.559,
      "step": 3630
    },
    {
      "epoch": 0.319915626647917,
      "grad_norm": 0.3579804599285126,
      "learning_rate": 2e-05,
      "loss": 0.5228,
      "step": 3640
    },
    {
      "epoch": 0.3207945157321146,
      "grad_norm": 0.36827799677848816,
      "learning_rate": 2e-05,
      "loss": 0.5861,
      "step": 3650
    },
    {
      "epoch": 0.3216734048163122,
      "grad_norm": 0.37947359681129456,
      "learning_rate": 2e-05,
      "loss": 0.5403,
      "step": 3660
    },
    {
      "epoch": 0.32255229390050977,
      "grad_norm": 0.34253841638565063,
      "learning_rate": 2e-05,
      "loss": 0.5205,
      "step": 3670
    },
    {
      "epoch": 0.32343118298470735,
      "grad_norm": 0.37916991114616394,
      "learning_rate": 2e-05,
      "loss": 0.5573,
      "step": 3680
    },
    {
      "epoch": 0.3243100720689049,
      "grad_norm": 0.3785051703453064,
      "learning_rate": 2e-05,
      "loss": 0.5497,
      "step": 3690
    },
    {
      "epoch": 0.3251889611531025,
      "grad_norm": 0.36646854877471924,
      "learning_rate": 2e-05,
      "loss": 0.5681,
      "step": 3700
    },
    {
      "epoch": 0.32606785023730006,
      "grad_norm": 0.3919835090637207,
      "learning_rate": 2e-05,
      "loss": 0.5499,
      "step": 3710
    },
    {
      "epoch": 0.32694673932149765,
      "grad_norm": 0.36537960171699524,
      "learning_rate": 2e-05,
      "loss": 0.5377,
      "step": 3720
    },
    {
      "epoch": 0.3278256284056952,
      "grad_norm": 0.3292663097381592,
      "learning_rate": 2e-05,
      "loss": 0.5147,
      "step": 3730
    },
    {
      "epoch": 0.32870451748989277,
      "grad_norm": 0.5153776407241821,
      "learning_rate": 2e-05,
      "loss": 0.5659,
      "step": 3740
    },
    {
      "epoch": 0.32958340657409035,
      "grad_norm": 0.3422275483608246,
      "learning_rate": 2e-05,
      "loss": 0.534,
      "step": 3750
    },
    {
      "epoch": 0.33046229565828794,
      "grad_norm": 0.3704967498779297,
      "learning_rate": 2e-05,
      "loss": 0.5554,
      "step": 3760
    },
    {
      "epoch": 0.3313411847424855,
      "grad_norm": 0.3404252827167511,
      "learning_rate": 2e-05,
      "loss": 0.5522,
      "step": 3770
    },
    {
      "epoch": 0.33222007382668306,
      "grad_norm": 0.3768712878227234,
      "learning_rate": 2e-05,
      "loss": 0.5656,
      "step": 3780
    },
    {
      "epoch": 0.33309896291088065,
      "grad_norm": 0.3620288670063019,
      "learning_rate": 2e-05,
      "loss": 0.5426,
      "step": 3790
    },
    {
      "epoch": 0.33397785199507823,
      "grad_norm": 0.3485812246799469,
      "learning_rate": 2e-05,
      "loss": 0.5649,
      "step": 3800
    },
    {
      "epoch": 0.3348567410792758,
      "grad_norm": 0.3773175776004791,
      "learning_rate": 2e-05,
      "loss": 0.5354,
      "step": 3810
    },
    {
      "epoch": 0.33573563016347335,
      "grad_norm": 0.3337115943431854,
      "learning_rate": 2e-05,
      "loss": 0.5237,
      "step": 3820
    },
    {
      "epoch": 0.33661451924767094,
      "grad_norm": 0.3432636559009552,
      "learning_rate": 2e-05,
      "loss": 0.5472,
      "step": 3830
    },
    {
      "epoch": 0.3374934083318685,
      "grad_norm": 0.34719574451446533,
      "learning_rate": 2e-05,
      "loss": 0.5549,
      "step": 3840
    },
    {
      "epoch": 0.3383722974160661,
      "grad_norm": 0.4003327786922455,
      "learning_rate": 2e-05,
      "loss": 0.5552,
      "step": 3850
    },
    {
      "epoch": 0.33925118650026365,
      "grad_norm": 0.33857205510139465,
      "learning_rate": 2e-05,
      "loss": 0.5514,
      "step": 3860
    },
    {
      "epoch": 0.34013007558446123,
      "grad_norm": 0.39958328008651733,
      "learning_rate": 2e-05,
      "loss": 0.5478,
      "step": 3870
    },
    {
      "epoch": 0.3410089646686588,
      "grad_norm": 0.3632514178752899,
      "learning_rate": 2e-05,
      "loss": 0.5475,
      "step": 3880
    },
    {
      "epoch": 0.3418878537528564,
      "grad_norm": 0.3496982157230377,
      "learning_rate": 2e-05,
      "loss": 0.5664,
      "step": 3890
    },
    {
      "epoch": 0.34276674283705394,
      "grad_norm": 0.36692944169044495,
      "learning_rate": 2e-05,
      "loss": 0.5593,
      "step": 3900
    },
    {
      "epoch": 0.3436456319212515,
      "grad_norm": 0.36224329471588135,
      "learning_rate": 2e-05,
      "loss": 0.5545,
      "step": 3910
    },
    {
      "epoch": 0.3445245210054491,
      "grad_norm": 0.3247584402561188,
      "learning_rate": 2e-05,
      "loss": 0.5464,
      "step": 3920
    },
    {
      "epoch": 0.3454034100896467,
      "grad_norm": 0.3404799997806549,
      "learning_rate": 2e-05,
      "loss": 0.5406,
      "step": 3930
    },
    {
      "epoch": 0.3462822991738443,
      "grad_norm": 0.3565092384815216,
      "learning_rate": 2e-05,
      "loss": 0.5262,
      "step": 3940
    },
    {
      "epoch": 0.3471611882580418,
      "grad_norm": 0.379717081785202,
      "learning_rate": 2e-05,
      "loss": 0.5366,
      "step": 3950
    },
    {
      "epoch": 0.3480400773422394,
      "grad_norm": 0.3445590138435364,
      "learning_rate": 2e-05,
      "loss": 0.5446,
      "step": 3960
    },
    {
      "epoch": 0.348918966426437,
      "grad_norm": 0.38638412952423096,
      "learning_rate": 2e-05,
      "loss": 0.5126,
      "step": 3970
    },
    {
      "epoch": 0.3497978555106346,
      "grad_norm": 0.3413693308830261,
      "learning_rate": 2e-05,
      "loss": 0.529,
      "step": 3980
    },
    {
      "epoch": 0.3506767445948321,
      "grad_norm": 0.35638323426246643,
      "learning_rate": 2e-05,
      "loss": 0.5663,
      "step": 3990
    },
    {
      "epoch": 0.3515556336790297,
      "grad_norm": 0.34048929810523987,
      "learning_rate": 2e-05,
      "loss": 0.524,
      "step": 4000
    },
    {
      "epoch": 0.3524345227632273,
      "grad_norm": 0.40266934037208557,
      "learning_rate": 2e-05,
      "loss": 0.5393,
      "step": 4010
    },
    {
      "epoch": 0.3533134118474249,
      "grad_norm": 0.3910423815250397,
      "learning_rate": 2e-05,
      "loss": 0.5617,
      "step": 4020
    },
    {
      "epoch": 0.3541923009316224,
      "grad_norm": 0.39282137155532837,
      "learning_rate": 2e-05,
      "loss": 0.5575,
      "step": 4030
    },
    {
      "epoch": 0.35507119001582,
      "grad_norm": 0.36524611711502075,
      "learning_rate": 2e-05,
      "loss": 0.559,
      "step": 4040
    },
    {
      "epoch": 0.3559500791000176,
      "grad_norm": 0.3243862986564636,
      "learning_rate": 2e-05,
      "loss": 0.5457,
      "step": 4050
    },
    {
      "epoch": 0.35682896818421517,
      "grad_norm": 0.31815096735954285,
      "learning_rate": 2e-05,
      "loss": 0.5532,
      "step": 4060
    },
    {
      "epoch": 0.3577078572684127,
      "grad_norm": 0.3698352873325348,
      "learning_rate": 2e-05,
      "loss": 0.5708,
      "step": 4070
    },
    {
      "epoch": 0.3585867463526103,
      "grad_norm": 0.4162037968635559,
      "learning_rate": 2e-05,
      "loss": 0.5794,
      "step": 4080
    },
    {
      "epoch": 0.3594656354368079,
      "grad_norm": 0.41902992129325867,
      "learning_rate": 2e-05,
      "loss": 0.565,
      "step": 4090
    },
    {
      "epoch": 0.36034452452100546,
      "grad_norm": 0.36418578028678894,
      "learning_rate": 2e-05,
      "loss": 0.5609,
      "step": 4100
    },
    {
      "epoch": 0.36122341360520305,
      "grad_norm": 0.4600087106227875,
      "learning_rate": 2e-05,
      "loss": 0.5629,
      "step": 4110
    },
    {
      "epoch": 0.3621023026894006,
      "grad_norm": 0.41888490319252014,
      "learning_rate": 2e-05,
      "loss": 0.5504,
      "step": 4120
    },
    {
      "epoch": 0.36298119177359817,
      "grad_norm": 0.36782190203666687,
      "learning_rate": 2e-05,
      "loss": 0.5252,
      "step": 4130
    },
    {
      "epoch": 0.36386008085779575,
      "grad_norm": 0.3366917073726654,
      "learning_rate": 2e-05,
      "loss": 0.5605,
      "step": 4140
    },
    {
      "epoch": 0.36473896994199334,
      "grad_norm": 0.3785319924354553,
      "learning_rate": 2e-05,
      "loss": 0.5632,
      "step": 4150
    },
    {
      "epoch": 0.3656178590261909,
      "grad_norm": 0.32834818959236145,
      "learning_rate": 2e-05,
      "loss": 0.5234,
      "step": 4160
    },
    {
      "epoch": 0.36649674811038846,
      "grad_norm": 0.3661729395389557,
      "learning_rate": 2e-05,
      "loss": 0.5687,
      "step": 4170
    },
    {
      "epoch": 0.36737563719458605,
      "grad_norm": 0.370278924703598,
      "learning_rate": 2e-05,
      "loss": 0.5517,
      "step": 4180
    },
    {
      "epoch": 0.36825452627878363,
      "grad_norm": 0.3554653227329254,
      "learning_rate": 2e-05,
      "loss": 0.5337,
      "step": 4190
    },
    {
      "epoch": 0.36913341536298117,
      "grad_norm": 0.35374918580055237,
      "learning_rate": 2e-05,
      "loss": 0.5384,
      "step": 4200
    },
    {
      "epoch": 0.37001230444717875,
      "grad_norm": 0.3210890591144562,
      "learning_rate": 2e-05,
      "loss": 0.5435,
      "step": 4210
    },
    {
      "epoch": 0.37089119353137634,
      "grad_norm": 0.3437519967556,
      "learning_rate": 2e-05,
      "loss": 0.5242,
      "step": 4220
    },
    {
      "epoch": 0.3717700826155739,
      "grad_norm": 0.35392579436302185,
      "learning_rate": 2e-05,
      "loss": 0.5472,
      "step": 4230
    },
    {
      "epoch": 0.3726489716997715,
      "grad_norm": 0.3219797611236572,
      "learning_rate": 2e-05,
      "loss": 0.5364,
      "step": 4240
    },
    {
      "epoch": 0.37352786078396905,
      "grad_norm": 0.36878517270088196,
      "learning_rate": 2e-05,
      "loss": 0.5483,
      "step": 4250
    },
    {
      "epoch": 0.37440674986816663,
      "grad_norm": 0.3591812551021576,
      "learning_rate": 2e-05,
      "loss": 0.5786,
      "step": 4260
    },
    {
      "epoch": 0.3752856389523642,
      "grad_norm": 0.337785542011261,
      "learning_rate": 2e-05,
      "loss": 0.5284,
      "step": 4270
    },
    {
      "epoch": 0.3761645280365618,
      "grad_norm": 0.38380974531173706,
      "learning_rate": 2e-05,
      "loss": 0.5532,
      "step": 4280
    },
    {
      "epoch": 0.37704341712075934,
      "grad_norm": 0.42100170254707336,
      "learning_rate": 2e-05,
      "loss": 0.5508,
      "step": 4290
    },
    {
      "epoch": 0.3779223062049569,
      "grad_norm": 0.335602343082428,
      "learning_rate": 2e-05,
      "loss": 0.5445,
      "step": 4300
    },
    {
      "epoch": 0.3788011952891545,
      "grad_norm": 0.341927170753479,
      "learning_rate": 2e-05,
      "loss": 0.5527,
      "step": 4310
    },
    {
      "epoch": 0.3796800843733521,
      "grad_norm": 0.3622886538505554,
      "learning_rate": 2e-05,
      "loss": 0.5381,
      "step": 4320
    },
    {
      "epoch": 0.38055897345754963,
      "grad_norm": 0.3521440029144287,
      "learning_rate": 2e-05,
      "loss": 0.5686,
      "step": 4330
    },
    {
      "epoch": 0.3814378625417472,
      "grad_norm": 0.34552690386772156,
      "learning_rate": 2e-05,
      "loss": 0.5529,
      "step": 4340
    },
    {
      "epoch": 0.3823167516259448,
      "grad_norm": 0.3347887098789215,
      "learning_rate": 2e-05,
      "loss": 0.5459,
      "step": 4350
    },
    {
      "epoch": 0.3831956407101424,
      "grad_norm": 0.341581791639328,
      "learning_rate": 2e-05,
      "loss": 0.5374,
      "step": 4360
    },
    {
      "epoch": 0.38407452979434,
      "grad_norm": 0.4854869544506073,
      "learning_rate": 2e-05,
      "loss": 0.5352,
      "step": 4370
    },
    {
      "epoch": 0.3849534188785375,
      "grad_norm": 0.3230288028717041,
      "learning_rate": 2e-05,
      "loss": 0.5492,
      "step": 4380
    },
    {
      "epoch": 0.3858323079627351,
      "grad_norm": 0.36125731468200684,
      "learning_rate": 2e-05,
      "loss": 0.5641,
      "step": 4390
    },
    {
      "epoch": 0.3867111970469327,
      "grad_norm": 0.30713826417922974,
      "learning_rate": 2e-05,
      "loss": 0.5484,
      "step": 4400
    },
    {
      "epoch": 0.3875900861311303,
      "grad_norm": 0.31878671050071716,
      "learning_rate": 2e-05,
      "loss": 0.5438,
      "step": 4410
    },
    {
      "epoch": 0.3884689752153278,
      "grad_norm": 0.38744089007377625,
      "learning_rate": 2e-05,
      "loss": 0.5309,
      "step": 4420
    },
    {
      "epoch": 0.3893478642995254,
      "grad_norm": 0.39923980832099915,
      "learning_rate": 2e-05,
      "loss": 0.5086,
      "step": 4430
    },
    {
      "epoch": 0.390226753383723,
      "grad_norm": 0.44628486037254333,
      "learning_rate": 2e-05,
      "loss": 0.5359,
      "step": 4440
    },
    {
      "epoch": 0.39110564246792057,
      "grad_norm": 0.3254205286502838,
      "learning_rate": 2e-05,
      "loss": 0.5181,
      "step": 4450
    },
    {
      "epoch": 0.3919845315521181,
      "grad_norm": 0.32878732681274414,
      "learning_rate": 2e-05,
      "loss": 0.5055,
      "step": 4460
    },
    {
      "epoch": 0.3928634206363157,
      "grad_norm": 0.35472363233566284,
      "learning_rate": 2e-05,
      "loss": 0.5707,
      "step": 4470
    },
    {
      "epoch": 0.3937423097205133,
      "grad_norm": 0.3601676821708679,
      "learning_rate": 2e-05,
      "loss": 0.5646,
      "step": 4480
    },
    {
      "epoch": 0.39462119880471086,
      "grad_norm": 0.3487088084220886,
      "learning_rate": 2e-05,
      "loss": 0.5271,
      "step": 4490
    },
    {
      "epoch": 0.39550008788890845,
      "grad_norm": 0.3668665885925293,
      "learning_rate": 2e-05,
      "loss": 0.5511,
      "step": 4500
    },
    {
      "epoch": 0.396378976973106,
      "grad_norm": 0.3314940333366394,
      "learning_rate": 2e-05,
      "loss": 0.5465,
      "step": 4510
    },
    {
      "epoch": 0.39725786605730357,
      "grad_norm": 0.3193863332271576,
      "learning_rate": 2e-05,
      "loss": 0.5265,
      "step": 4520
    },
    {
      "epoch": 0.39813675514150115,
      "grad_norm": 0.40486985445022583,
      "learning_rate": 2e-05,
      "loss": 0.5503,
      "step": 4530
    },
    {
      "epoch": 0.39901564422569874,
      "grad_norm": 0.44396767020225525,
      "learning_rate": 2e-05,
      "loss": 0.5451,
      "step": 4540
    },
    {
      "epoch": 0.39989453330989627,
      "grad_norm": 0.3681359589099884,
      "learning_rate": 2e-05,
      "loss": 0.5136,
      "step": 4550
    },
    {
      "epoch": 0.40077342239409386,
      "grad_norm": 0.35657134652137756,
      "learning_rate": 2e-05,
      "loss": 0.5721,
      "step": 4560
    },
    {
      "epoch": 0.40165231147829145,
      "grad_norm": 0.32666659355163574,
      "learning_rate": 2e-05,
      "loss": 0.5694,
      "step": 4570
    },
    {
      "epoch": 0.40253120056248903,
      "grad_norm": 0.3304166793823242,
      "learning_rate": 2e-05,
      "loss": 0.5437,
      "step": 4580
    },
    {
      "epoch": 0.40341008964668656,
      "grad_norm": 0.3637726604938507,
      "learning_rate": 2e-05,
      "loss": 0.5482,
      "step": 4590
    },
    {
      "epoch": 0.40428897873088415,
      "grad_norm": 0.4018683433532715,
      "learning_rate": 2e-05,
      "loss": 0.5307,
      "step": 4600
    },
    {
      "epoch": 0.40516786781508174,
      "grad_norm": 0.38370928168296814,
      "learning_rate": 2e-05,
      "loss": 0.5345,
      "step": 4610
    },
    {
      "epoch": 0.4060467568992793,
      "grad_norm": 0.40193265676498413,
      "learning_rate": 2e-05,
      "loss": 0.5574,
      "step": 4620
    },
    {
      "epoch": 0.40692564598347686,
      "grad_norm": 0.35073745250701904,
      "learning_rate": 2e-05,
      "loss": 0.544,
      "step": 4630
    },
    {
      "epoch": 0.40780453506767445,
      "grad_norm": 0.3612351715564728,
      "learning_rate": 2e-05,
      "loss": 0.5297,
      "step": 4640
    },
    {
      "epoch": 0.40868342415187203,
      "grad_norm": 0.3514283001422882,
      "learning_rate": 2e-05,
      "loss": 0.5577,
      "step": 4650
    },
    {
      "epoch": 0.4095623132360696,
      "grad_norm": 0.3579213619232178,
      "learning_rate": 2e-05,
      "loss": 0.5274,
      "step": 4660
    },
    {
      "epoch": 0.4104412023202672,
      "grad_norm": 0.3235413432121277,
      "learning_rate": 2e-05,
      "loss": 0.5421,
      "step": 4670
    },
    {
      "epoch": 0.41132009140446474,
      "grad_norm": 0.3696593642234802,
      "learning_rate": 2e-05,
      "loss": 0.5365,
      "step": 4680
    },
    {
      "epoch": 0.4121989804886623,
      "grad_norm": 0.37844300270080566,
      "learning_rate": 2e-05,
      "loss": 0.5359,
      "step": 4690
    },
    {
      "epoch": 0.4130778695728599,
      "grad_norm": 0.3614518940448761,
      "learning_rate": 2e-05,
      "loss": 0.5325,
      "step": 4700
    },
    {
      "epoch": 0.4139567586570575,
      "grad_norm": 0.35735979676246643,
      "learning_rate": 2e-05,
      "loss": 0.5411,
      "step": 4710
    },
    {
      "epoch": 0.41483564774125503,
      "grad_norm": 0.3506315052509308,
      "learning_rate": 2e-05,
      "loss": 0.5416,
      "step": 4720
    },
    {
      "epoch": 0.4157145368254526,
      "grad_norm": 0.33660030364990234,
      "learning_rate": 2e-05,
      "loss": 0.5641,
      "step": 4730
    },
    {
      "epoch": 0.4165934259096502,
      "grad_norm": 0.3599470555782318,
      "learning_rate": 2e-05,
      "loss": 0.555,
      "step": 4740
    },
    {
      "epoch": 0.4174723149938478,
      "grad_norm": 0.34064146876335144,
      "learning_rate": 2e-05,
      "loss": 0.5394,
      "step": 4750
    },
    {
      "epoch": 0.4183512040780453,
      "grad_norm": 0.39736467599868774,
      "learning_rate": 2e-05,
      "loss": 0.5514,
      "step": 4760
    },
    {
      "epoch": 0.4192300931622429,
      "grad_norm": 0.34742310643196106,
      "learning_rate": 2e-05,
      "loss": 0.5481,
      "step": 4770
    },
    {
      "epoch": 0.4201089822464405,
      "grad_norm": 0.39449450373649597,
      "learning_rate": 2e-05,
      "loss": 0.5515,
      "step": 4780
    },
    {
      "epoch": 0.4209878713306381,
      "grad_norm": 0.46250203251838684,
      "learning_rate": 2e-05,
      "loss": 0.5621,
      "step": 4790
    },
    {
      "epoch": 0.4218667604148357,
      "grad_norm": 0.4337044060230255,
      "learning_rate": 2e-05,
      "loss": 0.5538,
      "step": 4800
    },
    {
      "epoch": 0.4227456494990332,
      "grad_norm": 0.4036795496940613,
      "learning_rate": 2e-05,
      "loss": 0.5553,
      "step": 4810
    },
    {
      "epoch": 0.4236245385832308,
      "grad_norm": 0.34802335500717163,
      "learning_rate": 2e-05,
      "loss": 0.5426,
      "step": 4820
    },
    {
      "epoch": 0.4245034276674284,
      "grad_norm": 0.3582998812198639,
      "learning_rate": 2e-05,
      "loss": 0.5545,
      "step": 4830
    },
    {
      "epoch": 0.42538231675162597,
      "grad_norm": 0.38593223690986633,
      "learning_rate": 2e-05,
      "loss": 0.5508,
      "step": 4840
    },
    {
      "epoch": 0.4262612058358235,
      "grad_norm": 0.33207061886787415,
      "learning_rate": 2e-05,
      "loss": 0.5376,
      "step": 4850
    },
    {
      "epoch": 0.4271400949200211,
      "grad_norm": 0.3460117280483246,
      "learning_rate": 2e-05,
      "loss": 0.5302,
      "step": 4860
    },
    {
      "epoch": 0.42801898400421867,
      "grad_norm": 0.35605132579803467,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 4870
    },
    {
      "epoch": 0.42889787308841626,
      "grad_norm": 0.34891632199287415,
      "learning_rate": 2e-05,
      "loss": 0.5504,
      "step": 4880
    },
    {
      "epoch": 0.4297767621726138,
      "grad_norm": 0.41499564051628113,
      "learning_rate": 2e-05,
      "loss": 0.5672,
      "step": 4890
    },
    {
      "epoch": 0.4306556512568114,
      "grad_norm": 0.37195125222206116,
      "learning_rate": 2e-05,
      "loss": 0.5288,
      "step": 4900
    },
    {
      "epoch": 0.43153454034100897,
      "grad_norm": 0.32856228947639465,
      "learning_rate": 2e-05,
      "loss": 0.5477,
      "step": 4910
    },
    {
      "epoch": 0.43241342942520655,
      "grad_norm": 0.3535173833370209,
      "learning_rate": 2e-05,
      "loss": 0.5398,
      "step": 4920
    },
    {
      "epoch": 0.43329231850940414,
      "grad_norm": 0.3696601986885071,
      "learning_rate": 2e-05,
      "loss": 0.541,
      "step": 4930
    },
    {
      "epoch": 0.43417120759360167,
      "grad_norm": 0.379461407661438,
      "learning_rate": 2e-05,
      "loss": 0.5546,
      "step": 4940
    },
    {
      "epoch": 0.43505009667779926,
      "grad_norm": 0.36268851161003113,
      "learning_rate": 2e-05,
      "loss": 0.5456,
      "step": 4950
    },
    {
      "epoch": 0.43592898576199685,
      "grad_norm": 0.4394989609718323,
      "learning_rate": 2e-05,
      "loss": 0.5368,
      "step": 4960
    },
    {
      "epoch": 0.43680787484619443,
      "grad_norm": 0.35105985403060913,
      "learning_rate": 2e-05,
      "loss": 0.5086,
      "step": 4970
    },
    {
      "epoch": 0.43768676393039196,
      "grad_norm": 0.37677186727523804,
      "learning_rate": 2e-05,
      "loss": 0.5373,
      "step": 4980
    },
    {
      "epoch": 0.43856565301458955,
      "grad_norm": 0.3612132668495178,
      "learning_rate": 2e-05,
      "loss": 0.5389,
      "step": 4990
    },
    {
      "epoch": 0.43944454209878714,
      "grad_norm": 0.3109592795372009,
      "learning_rate": 2e-05,
      "loss": 0.5237,
      "step": 5000
    },
    {
      "epoch": 0.4403234311829847,
      "grad_norm": 0.42327675223350525,
      "learning_rate": 2e-05,
      "loss": 0.5209,
      "step": 5010
    },
    {
      "epoch": 0.44120232026718226,
      "grad_norm": 0.32862475514411926,
      "learning_rate": 2e-05,
      "loss": 0.5364,
      "step": 5020
    },
    {
      "epoch": 0.44208120935137984,
      "grad_norm": 0.3787497878074646,
      "learning_rate": 2e-05,
      "loss": 0.5379,
      "step": 5030
    },
    {
      "epoch": 0.44296009843557743,
      "grad_norm": 0.4235406816005707,
      "learning_rate": 2e-05,
      "loss": 0.5118,
      "step": 5040
    },
    {
      "epoch": 0.443838987519775,
      "grad_norm": 0.34716659784317017,
      "learning_rate": 2e-05,
      "loss": 0.5367,
      "step": 5050
    },
    {
      "epoch": 0.4447178766039726,
      "grad_norm": 0.3423392176628113,
      "learning_rate": 2e-05,
      "loss": 0.5378,
      "step": 5060
    },
    {
      "epoch": 0.44559676568817014,
      "grad_norm": 0.36973053216934204,
      "learning_rate": 2e-05,
      "loss": 0.5516,
      "step": 5070
    },
    {
      "epoch": 0.4464756547723677,
      "grad_norm": 0.4135674238204956,
      "learning_rate": 2e-05,
      "loss": 0.5289,
      "step": 5080
    },
    {
      "epoch": 0.4473545438565653,
      "grad_norm": 0.34592947363853455,
      "learning_rate": 2e-05,
      "loss": 0.589,
      "step": 5090
    },
    {
      "epoch": 0.4482334329407629,
      "grad_norm": 0.361943781375885,
      "learning_rate": 2e-05,
      "loss": 0.5323,
      "step": 5100
    },
    {
      "epoch": 0.44911232202496043,
      "grad_norm": 0.33657971024513245,
      "learning_rate": 2e-05,
      "loss": 0.5446,
      "step": 5110
    },
    {
      "epoch": 0.449991211109158,
      "grad_norm": 0.40074422955513,
      "learning_rate": 2e-05,
      "loss": 0.5169,
      "step": 5120
    },
    {
      "epoch": 0.4508701001933556,
      "grad_norm": 0.382124662399292,
      "learning_rate": 2e-05,
      "loss": 0.5261,
      "step": 5130
    },
    {
      "epoch": 0.4517489892775532,
      "grad_norm": 0.3303861916065216,
      "learning_rate": 2e-05,
      "loss": 0.5403,
      "step": 5140
    },
    {
      "epoch": 0.4526278783617507,
      "grad_norm": 0.4123605489730835,
      "learning_rate": 2e-05,
      "loss": 0.527,
      "step": 5150
    },
    {
      "epoch": 0.4535067674459483,
      "grad_norm": 0.3331770896911621,
      "learning_rate": 2e-05,
      "loss": 0.5415,
      "step": 5160
    },
    {
      "epoch": 0.4543856565301459,
      "grad_norm": 0.3451739549636841,
      "learning_rate": 2e-05,
      "loss": 0.5407,
      "step": 5170
    },
    {
      "epoch": 0.4552645456143435,
      "grad_norm": 0.4306280016899109,
      "learning_rate": 2e-05,
      "loss": 0.5282,
      "step": 5180
    },
    {
      "epoch": 0.456143434698541,
      "grad_norm": 0.32369932532310486,
      "learning_rate": 2e-05,
      "loss": 0.5484,
      "step": 5190
    },
    {
      "epoch": 0.4570223237827386,
      "grad_norm": 0.33302465081214905,
      "learning_rate": 2e-05,
      "loss": 0.5156,
      "step": 5200
    },
    {
      "epoch": 0.4579012128669362,
      "grad_norm": 0.3375796973705292,
      "learning_rate": 2e-05,
      "loss": 0.5272,
      "step": 5210
    },
    {
      "epoch": 0.4587801019511338,
      "grad_norm": 0.35944458842277527,
      "learning_rate": 2e-05,
      "loss": 0.5172,
      "step": 5220
    },
    {
      "epoch": 0.45965899103533137,
      "grad_norm": 0.35269901156425476,
      "learning_rate": 2e-05,
      "loss": 0.5324,
      "step": 5230
    },
    {
      "epoch": 0.4605378801195289,
      "grad_norm": 0.3802284896373749,
      "learning_rate": 2e-05,
      "loss": 0.5361,
      "step": 5240
    },
    {
      "epoch": 0.4614167692037265,
      "grad_norm": 0.34590405225753784,
      "learning_rate": 2e-05,
      "loss": 0.5473,
      "step": 5250
    },
    {
      "epoch": 0.46229565828792407,
      "grad_norm": 0.35689690709114075,
      "learning_rate": 2e-05,
      "loss": 0.5446,
      "step": 5260
    },
    {
      "epoch": 0.46317454737212166,
      "grad_norm": 0.34083884954452515,
      "learning_rate": 2e-05,
      "loss": 0.5728,
      "step": 5270
    },
    {
      "epoch": 0.4640534364563192,
      "grad_norm": 0.3582720458507538,
      "learning_rate": 2e-05,
      "loss": 0.549,
      "step": 5280
    },
    {
      "epoch": 0.4649323255405168,
      "grad_norm": 0.3676546812057495,
      "learning_rate": 2e-05,
      "loss": 0.526,
      "step": 5290
    },
    {
      "epoch": 0.46581121462471436,
      "grad_norm": 0.36298972368240356,
      "learning_rate": 2e-05,
      "loss": 0.5312,
      "step": 5300
    },
    {
      "epoch": 0.46669010370891195,
      "grad_norm": 0.3767959475517273,
      "learning_rate": 2e-05,
      "loss": 0.5204,
      "step": 5310
    },
    {
      "epoch": 0.4675689927931095,
      "grad_norm": 0.35869765281677246,
      "learning_rate": 2e-05,
      "loss": 0.5266,
      "step": 5320
    },
    {
      "epoch": 0.46844788187730707,
      "grad_norm": 0.36787673830986023,
      "learning_rate": 2e-05,
      "loss": 0.5638,
      "step": 5330
    },
    {
      "epoch": 0.46932677096150466,
      "grad_norm": 0.38523900508880615,
      "learning_rate": 2e-05,
      "loss": 0.5573,
      "step": 5340
    },
    {
      "epoch": 0.47020566004570225,
      "grad_norm": 0.3652077615261078,
      "learning_rate": 2e-05,
      "loss": 0.5414,
      "step": 5350
    },
    {
      "epoch": 0.47108454912989983,
      "grad_norm": 0.3905850052833557,
      "learning_rate": 2e-05,
      "loss": 0.549,
      "step": 5360
    },
    {
      "epoch": 0.47196343821409736,
      "grad_norm": 0.37188977003097534,
      "learning_rate": 2e-05,
      "loss": 0.5344,
      "step": 5370
    },
    {
      "epoch": 0.47284232729829495,
      "grad_norm": 0.33724963665008545,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 5380
    },
    {
      "epoch": 0.47372121638249254,
      "grad_norm": 0.37381434440612793,
      "learning_rate": 2e-05,
      "loss": 0.5246,
      "step": 5390
    },
    {
      "epoch": 0.4746001054666901,
      "grad_norm": 0.3596455454826355,
      "learning_rate": 2e-05,
      "loss": 0.5271,
      "step": 5400
    },
    {
      "epoch": 0.47547899455088766,
      "grad_norm": 0.3414129316806793,
      "learning_rate": 2e-05,
      "loss": 0.526,
      "step": 5410
    },
    {
      "epoch": 0.47635788363508524,
      "grad_norm": 0.3871542811393738,
      "learning_rate": 2e-05,
      "loss": 0.5394,
      "step": 5420
    },
    {
      "epoch": 0.47723677271928283,
      "grad_norm": 0.6655234694480896,
      "learning_rate": 2e-05,
      "loss": 0.5466,
      "step": 5430
    },
    {
      "epoch": 0.4781156618034804,
      "grad_norm": 0.3557099997997284,
      "learning_rate": 2e-05,
      "loss": 0.5243,
      "step": 5440
    },
    {
      "epoch": 0.47899455088767795,
      "grad_norm": 0.3287471532821655,
      "learning_rate": 2e-05,
      "loss": 0.5416,
      "step": 5450
    },
    {
      "epoch": 0.47987343997187554,
      "grad_norm": 0.3387470543384552,
      "learning_rate": 2e-05,
      "loss": 0.5451,
      "step": 5460
    },
    {
      "epoch": 0.4807523290560731,
      "grad_norm": 0.341594934463501,
      "learning_rate": 2e-05,
      "loss": 0.5608,
      "step": 5470
    },
    {
      "epoch": 0.4816312181402707,
      "grad_norm": 0.3416372835636139,
      "learning_rate": 2e-05,
      "loss": 0.5244,
      "step": 5480
    },
    {
      "epoch": 0.4825101072244683,
      "grad_norm": 0.33682698011398315,
      "learning_rate": 2e-05,
      "loss": 0.5431,
      "step": 5490
    },
    {
      "epoch": 0.48338899630866583,
      "grad_norm": 0.3447750210762024,
      "learning_rate": 2e-05,
      "loss": 0.529,
      "step": 5500
    },
    {
      "epoch": 0.4842678853928634,
      "grad_norm": 0.3910992443561554,
      "learning_rate": 2e-05,
      "loss": 0.5457,
      "step": 5510
    },
    {
      "epoch": 0.485146774477061,
      "grad_norm": 0.3725154995918274,
      "learning_rate": 2e-05,
      "loss": 0.5482,
      "step": 5520
    },
    {
      "epoch": 0.4860256635612586,
      "grad_norm": 0.37032729387283325,
      "learning_rate": 2e-05,
      "loss": 0.5412,
      "step": 5530
    },
    {
      "epoch": 0.4869045526454561,
      "grad_norm": 0.42059218883514404,
      "learning_rate": 2e-05,
      "loss": 0.5281,
      "step": 5540
    },
    {
      "epoch": 0.4877834417296537,
      "grad_norm": 0.39963313937187195,
      "learning_rate": 2e-05,
      "loss": 0.5537,
      "step": 5550
    },
    {
      "epoch": 0.4886623308138513,
      "grad_norm": 0.34657004475593567,
      "learning_rate": 2e-05,
      "loss": 0.5512,
      "step": 5560
    },
    {
      "epoch": 0.4895412198980489,
      "grad_norm": 0.33465126156806946,
      "learning_rate": 2e-05,
      "loss": 0.5276,
      "step": 5570
    },
    {
      "epoch": 0.4904201089822464,
      "grad_norm": 0.3606172800064087,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 5580
    },
    {
      "epoch": 0.491298998066444,
      "grad_norm": 0.3252260088920593,
      "learning_rate": 2e-05,
      "loss": 0.5285,
      "step": 5590
    },
    {
      "epoch": 0.4921778871506416,
      "grad_norm": 0.35943230986595154,
      "learning_rate": 2e-05,
      "loss": 0.5515,
      "step": 5600
    },
    {
      "epoch": 0.4930567762348392,
      "grad_norm": 0.3700086772441864,
      "learning_rate": 2e-05,
      "loss": 0.5443,
      "step": 5610
    },
    {
      "epoch": 0.49393566531903677,
      "grad_norm": 0.3655369281768799,
      "learning_rate": 2e-05,
      "loss": 0.5551,
      "step": 5620
    },
    {
      "epoch": 0.4948145544032343,
      "grad_norm": 0.39791229367256165,
      "learning_rate": 2e-05,
      "loss": 0.5223,
      "step": 5630
    },
    {
      "epoch": 0.4956934434874319,
      "grad_norm": 0.32812586426734924,
      "learning_rate": 2e-05,
      "loss": 0.5368,
      "step": 5640
    },
    {
      "epoch": 0.49657233257162947,
      "grad_norm": 0.3422529101371765,
      "learning_rate": 2e-05,
      "loss": 0.5718,
      "step": 5650
    },
    {
      "epoch": 0.49745122165582706,
      "grad_norm": 0.3405756652355194,
      "learning_rate": 2e-05,
      "loss": 0.5621,
      "step": 5660
    },
    {
      "epoch": 0.4983301107400246,
      "grad_norm": 0.3696461319923401,
      "learning_rate": 2e-05,
      "loss": 0.5368,
      "step": 5670
    },
    {
      "epoch": 0.4992089998242222,
      "grad_norm": 0.38700246810913086,
      "learning_rate": 2e-05,
      "loss": 0.5315,
      "step": 5680
    },
    {
      "epoch": 0.5000878889084197,
      "grad_norm": 0.3487662672996521,
      "learning_rate": 2e-05,
      "loss": 0.5225,
      "step": 5690
    },
    {
      "epoch": 0.5009667779926174,
      "grad_norm": 0.3654389977455139,
      "learning_rate": 2e-05,
      "loss": 0.5428,
      "step": 5700
    },
    {
      "epoch": 0.5018456670768149,
      "grad_norm": 0.3368045687675476,
      "learning_rate": 2e-05,
      "loss": 0.5154,
      "step": 5710
    },
    {
      "epoch": 0.5027245561610125,
      "grad_norm": 0.4672960042953491,
      "learning_rate": 2e-05,
      "loss": 0.5586,
      "step": 5720
    },
    {
      "epoch": 0.5036034452452101,
      "grad_norm": 0.33891552686691284,
      "learning_rate": 2e-05,
      "loss": 0.531,
      "step": 5730
    },
    {
      "epoch": 0.5044823343294076,
      "grad_norm": 0.3513313829898834,
      "learning_rate": 2e-05,
      "loss": 0.5268,
      "step": 5740
    },
    {
      "epoch": 0.5053612234136052,
      "grad_norm": 0.3382588326931,
      "learning_rate": 2e-05,
      "loss": 0.5421,
      "step": 5750
    },
    {
      "epoch": 0.5062401124978028,
      "grad_norm": 0.36326006054878235,
      "learning_rate": 2e-05,
      "loss": 0.5403,
      "step": 5760
    },
    {
      "epoch": 0.5071190015820004,
      "grad_norm": 0.353162556886673,
      "learning_rate": 2e-05,
      "loss": 0.5458,
      "step": 5770
    },
    {
      "epoch": 0.5079978906661979,
      "grad_norm": 0.3601173460483551,
      "learning_rate": 2e-05,
      "loss": 0.5046,
      "step": 5780
    },
    {
      "epoch": 0.5088767797503955,
      "grad_norm": 0.36387717723846436,
      "learning_rate": 2e-05,
      "loss": 0.506,
      "step": 5790
    },
    {
      "epoch": 0.5097556688345931,
      "grad_norm": 2.5938162803649902,
      "learning_rate": 2e-05,
      "loss": 0.5371,
      "step": 5800
    },
    {
      "epoch": 0.5106345579187906,
      "grad_norm": 0.4095554053783417,
      "learning_rate": 2e-05,
      "loss": 0.545,
      "step": 5810
    },
    {
      "epoch": 0.5115134470029882,
      "grad_norm": 0.4038480520248413,
      "learning_rate": 2e-05,
      "loss": 0.5134,
      "step": 5820
    },
    {
      "epoch": 0.5123923360871858,
      "grad_norm": 0.3482191562652588,
      "learning_rate": 2e-05,
      "loss": 0.5361,
      "step": 5830
    },
    {
      "epoch": 0.5132712251713833,
      "grad_norm": 0.3524332046508789,
      "learning_rate": 2e-05,
      "loss": 0.5252,
      "step": 5840
    },
    {
      "epoch": 0.514150114255581,
      "grad_norm": 0.39291709661483765,
      "learning_rate": 2e-05,
      "loss": 0.5233,
      "step": 5850
    },
    {
      "epoch": 0.5150290033397785,
      "grad_norm": 0.4045007824897766,
      "learning_rate": 2e-05,
      "loss": 0.5531,
      "step": 5860
    },
    {
      "epoch": 0.5159078924239761,
      "grad_norm": 0.3543850779533386,
      "learning_rate": 2e-05,
      "loss": 0.5225,
      "step": 5870
    },
    {
      "epoch": 0.5167867815081737,
      "grad_norm": 0.4048401415348053,
      "learning_rate": 2e-05,
      "loss": 0.5383,
      "step": 5880
    },
    {
      "epoch": 0.5176656705923712,
      "grad_norm": 0.35716232657432556,
      "learning_rate": 2e-05,
      "loss": 0.5546,
      "step": 5890
    },
    {
      "epoch": 0.5185445596765689,
      "grad_norm": 0.3323361873626709,
      "learning_rate": 2e-05,
      "loss": 0.5305,
      "step": 5900
    },
    {
      "epoch": 0.5194234487607664,
      "grad_norm": 0.4494822025299072,
      "learning_rate": 2e-05,
      "loss": 0.5367,
      "step": 5910
    },
    {
      "epoch": 0.5203023378449639,
      "grad_norm": 0.37239599227905273,
      "learning_rate": 2e-05,
      "loss": 0.5177,
      "step": 5920
    },
    {
      "epoch": 0.5211812269291616,
      "grad_norm": 0.47401344776153564,
      "learning_rate": 2e-05,
      "loss": 0.5527,
      "step": 5930
    },
    {
      "epoch": 0.5220601160133591,
      "grad_norm": 0.38761821389198303,
      "learning_rate": 2e-05,
      "loss": 0.5113,
      "step": 5940
    },
    {
      "epoch": 0.5229390050975566,
      "grad_norm": 0.3818376958370209,
      "learning_rate": 2e-05,
      "loss": 0.5348,
      "step": 5950
    },
    {
      "epoch": 0.5238178941817543,
      "grad_norm": 0.3191867172718048,
      "learning_rate": 2e-05,
      "loss": 0.5519,
      "step": 5960
    },
    {
      "epoch": 0.5246967832659518,
      "grad_norm": 0.3737521767616272,
      "learning_rate": 2e-05,
      "loss": 0.5361,
      "step": 5970
    },
    {
      "epoch": 0.5255756723501495,
      "grad_norm": 0.3138112425804138,
      "learning_rate": 2e-05,
      "loss": 0.54,
      "step": 5980
    },
    {
      "epoch": 0.526454561434347,
      "grad_norm": 0.3004269301891327,
      "learning_rate": 2e-05,
      "loss": 0.5156,
      "step": 5990
    },
    {
      "epoch": 0.5273334505185445,
      "grad_norm": 0.363715797662735,
      "learning_rate": 2e-05,
      "loss": 0.5251,
      "step": 6000
    },
    {
      "epoch": 0.5282123396027422,
      "grad_norm": 0.3640502095222473,
      "learning_rate": 2e-05,
      "loss": 0.5173,
      "step": 6010
    },
    {
      "epoch": 0.5290912286869397,
      "grad_norm": 0.3088285028934479,
      "learning_rate": 2e-05,
      "loss": 0.5223,
      "step": 6020
    },
    {
      "epoch": 0.5299701177711372,
      "grad_norm": 0.3617301881313324,
      "learning_rate": 2e-05,
      "loss": 0.531,
      "step": 6030
    },
    {
      "epoch": 0.5308490068553349,
      "grad_norm": 0.3612990975379944,
      "learning_rate": 2e-05,
      "loss": 0.5453,
      "step": 6040
    },
    {
      "epoch": 0.5317278959395324,
      "grad_norm": 0.35616442561149597,
      "learning_rate": 2e-05,
      "loss": 0.5194,
      "step": 6050
    },
    {
      "epoch": 0.53260678502373,
      "grad_norm": 0.5063347220420837,
      "learning_rate": 2e-05,
      "loss": 0.5614,
      "step": 6060
    },
    {
      "epoch": 0.5334856741079276,
      "grad_norm": 0.37985149025917053,
      "learning_rate": 2e-05,
      "loss": 0.5179,
      "step": 6070
    },
    {
      "epoch": 0.5343645631921251,
      "grad_norm": 0.4794306457042694,
      "learning_rate": 2e-05,
      "loss": 0.5397,
      "step": 6080
    },
    {
      "epoch": 0.5352434522763228,
      "grad_norm": 0.33605316281318665,
      "learning_rate": 2e-05,
      "loss": 0.5219,
      "step": 6090
    },
    {
      "epoch": 0.5361223413605203,
      "grad_norm": 0.36132848262786865,
      "learning_rate": 2e-05,
      "loss": 0.5454,
      "step": 6100
    },
    {
      "epoch": 0.5370012304447179,
      "grad_norm": 0.39514991641044617,
      "learning_rate": 2e-05,
      "loss": 0.5518,
      "step": 6110
    },
    {
      "epoch": 0.5378801195289155,
      "grad_norm": 0.6602166891098022,
      "learning_rate": 2e-05,
      "loss": 0.5107,
      "step": 6120
    },
    {
      "epoch": 0.538759008613113,
      "grad_norm": 0.4659086763858795,
      "learning_rate": 2e-05,
      "loss": 0.5689,
      "step": 6130
    },
    {
      "epoch": 0.5396378976973106,
      "grad_norm": 0.33729246258735657,
      "learning_rate": 2e-05,
      "loss": 0.5359,
      "step": 6140
    },
    {
      "epoch": 0.5405167867815082,
      "grad_norm": 0.38420701026916504,
      "learning_rate": 2e-05,
      "loss": 0.5223,
      "step": 6150
    },
    {
      "epoch": 0.5413956758657057,
      "grad_norm": 0.3617807626724243,
      "learning_rate": 2e-05,
      "loss": 0.5346,
      "step": 6160
    },
    {
      "epoch": 0.5422745649499033,
      "grad_norm": 0.3897862136363983,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 6170
    },
    {
      "epoch": 0.5431534540341009,
      "grad_norm": 0.33242279291152954,
      "learning_rate": 2e-05,
      "loss": 0.5168,
      "step": 6180
    },
    {
      "epoch": 0.5440323431182985,
      "grad_norm": 0.35702192783355713,
      "learning_rate": 2e-05,
      "loss": 0.564,
      "step": 6190
    },
    {
      "epoch": 0.544911232202496,
      "grad_norm": 0.38272905349731445,
      "learning_rate": 2e-05,
      "loss": 0.5053,
      "step": 6200
    },
    {
      "epoch": 0.5457901212866936,
      "grad_norm": 0.41474470496177673,
      "learning_rate": 2e-05,
      "loss": 0.4909,
      "step": 6210
    },
    {
      "epoch": 0.5466690103708912,
      "grad_norm": 0.35466277599334717,
      "learning_rate": 2e-05,
      "loss": 0.5255,
      "step": 6220
    },
    {
      "epoch": 0.5475478994550887,
      "grad_norm": 0.3527427017688751,
      "learning_rate": 2e-05,
      "loss": 0.5167,
      "step": 6230
    },
    {
      "epoch": 0.5484267885392864,
      "grad_norm": 0.3785998225212097,
      "learning_rate": 2e-05,
      "loss": 0.5238,
      "step": 6240
    },
    {
      "epoch": 0.5493056776234839,
      "grad_norm": 0.36652424931526184,
      "learning_rate": 2e-05,
      "loss": 0.5306,
      "step": 6250
    },
    {
      "epoch": 0.5501845667076815,
      "grad_norm": 0.38643917441368103,
      "learning_rate": 2e-05,
      "loss": 0.5325,
      "step": 6260
    },
    {
      "epoch": 0.5510634557918791,
      "grad_norm": 0.3624691665172577,
      "learning_rate": 2e-05,
      "loss": 0.5302,
      "step": 6270
    },
    {
      "epoch": 0.5519423448760766,
      "grad_norm": 0.3505941927433014,
      "learning_rate": 2e-05,
      "loss": 0.5344,
      "step": 6280
    },
    {
      "epoch": 0.5528212339602742,
      "grad_norm": 0.38157919049263,
      "learning_rate": 2e-05,
      "loss": 0.509,
      "step": 6290
    },
    {
      "epoch": 0.5537001230444718,
      "grad_norm": 0.47855067253112793,
      "learning_rate": 2e-05,
      "loss": 0.5353,
      "step": 6300
    },
    {
      "epoch": 0.5545790121286693,
      "grad_norm": 0.3783285617828369,
      "learning_rate": 2e-05,
      "loss": 0.5113,
      "step": 6310
    },
    {
      "epoch": 0.555457901212867,
      "grad_norm": 0.39671871066093445,
      "learning_rate": 2e-05,
      "loss": 0.5323,
      "step": 6320
    },
    {
      "epoch": 0.5563367902970645,
      "grad_norm": 0.382790207862854,
      "learning_rate": 2e-05,
      "loss": 0.5245,
      "step": 6330
    },
    {
      "epoch": 0.557215679381262,
      "grad_norm": 0.3687851130962372,
      "learning_rate": 2e-05,
      "loss": 0.5186,
      "step": 6340
    },
    {
      "epoch": 0.5580945684654597,
      "grad_norm": 0.3768340051174164,
      "learning_rate": 2e-05,
      "loss": 0.5452,
      "step": 6350
    },
    {
      "epoch": 0.5589734575496572,
      "grad_norm": 0.380201518535614,
      "learning_rate": 2e-05,
      "loss": 0.4905,
      "step": 6360
    },
    {
      "epoch": 0.5598523466338549,
      "grad_norm": 0.36144688725471497,
      "learning_rate": 2e-05,
      "loss": 0.5313,
      "step": 6370
    },
    {
      "epoch": 0.5607312357180524,
      "grad_norm": 0.39105117321014404,
      "learning_rate": 2e-05,
      "loss": 0.5274,
      "step": 6380
    },
    {
      "epoch": 0.5616101248022499,
      "grad_norm": 0.34880003333091736,
      "learning_rate": 2e-05,
      "loss": 0.5204,
      "step": 6390
    },
    {
      "epoch": 0.5624890138864476,
      "grad_norm": 0.35628801584243774,
      "learning_rate": 2e-05,
      "loss": 0.5114,
      "step": 6400
    },
    {
      "epoch": 0.5633679029706451,
      "grad_norm": 0.32866501808166504,
      "learning_rate": 2e-05,
      "loss": 0.5369,
      "step": 6410
    },
    {
      "epoch": 0.5642467920548426,
      "grad_norm": 0.3618732690811157,
      "learning_rate": 2e-05,
      "loss": 0.5282,
      "step": 6420
    },
    {
      "epoch": 0.5651256811390403,
      "grad_norm": 0.35647842288017273,
      "learning_rate": 2e-05,
      "loss": 0.5138,
      "step": 6430
    },
    {
      "epoch": 0.5660045702232378,
      "grad_norm": 0.40110349655151367,
      "learning_rate": 2e-05,
      "loss": 0.532,
      "step": 6440
    },
    {
      "epoch": 0.5668834593074354,
      "grad_norm": 0.3437333405017853,
      "learning_rate": 2e-05,
      "loss": 0.5159,
      "step": 6450
    },
    {
      "epoch": 0.567762348391633,
      "grad_norm": 0.5862695574760437,
      "learning_rate": 2e-05,
      "loss": 0.5196,
      "step": 6460
    },
    {
      "epoch": 0.5686412374758305,
      "grad_norm": 0.32273033261299133,
      "learning_rate": 2e-05,
      "loss": 0.5183,
      "step": 6470
    },
    {
      "epoch": 0.5695201265600282,
      "grad_norm": 0.40969178080558777,
      "learning_rate": 2e-05,
      "loss": 0.5206,
      "step": 6480
    },
    {
      "epoch": 0.5703990156442257,
      "grad_norm": 0.33422744274139404,
      "learning_rate": 2e-05,
      "loss": 0.5409,
      "step": 6490
    },
    {
      "epoch": 0.5712779047284233,
      "grad_norm": 0.35028237104415894,
      "learning_rate": 2e-05,
      "loss": 0.5381,
      "step": 6500
    },
    {
      "epoch": 0.5721567938126209,
      "grad_norm": 0.3817231357097626,
      "learning_rate": 2e-05,
      "loss": 0.5094,
      "step": 6510
    },
    {
      "epoch": 0.5730356828968184,
      "grad_norm": 0.3767814636230469,
      "learning_rate": 2e-05,
      "loss": 0.5573,
      "step": 6520
    },
    {
      "epoch": 0.573914571981016,
      "grad_norm": 0.3713395297527313,
      "learning_rate": 2e-05,
      "loss": 0.5435,
      "step": 6530
    },
    {
      "epoch": 0.5747934610652136,
      "grad_norm": 0.3787616491317749,
      "learning_rate": 2e-05,
      "loss": 0.5523,
      "step": 6540
    },
    {
      "epoch": 0.5756723501494111,
      "grad_norm": 0.4079139530658722,
      "learning_rate": 2e-05,
      "loss": 0.5474,
      "step": 6550
    },
    {
      "epoch": 0.5765512392336087,
      "grad_norm": 0.32201075553894043,
      "learning_rate": 2e-05,
      "loss": 0.5273,
      "step": 6560
    },
    {
      "epoch": 0.5774301283178063,
      "grad_norm": 0.34729668498039246,
      "learning_rate": 2e-05,
      "loss": 0.5115,
      "step": 6570
    },
    {
      "epoch": 0.5783090174020039,
      "grad_norm": 0.3866352438926697,
      "learning_rate": 2e-05,
      "loss": 0.5402,
      "step": 6580
    },
    {
      "epoch": 0.5791879064862014,
      "grad_norm": 0.3747514486312866,
      "learning_rate": 2e-05,
      "loss": 0.5338,
      "step": 6590
    },
    {
      "epoch": 0.580066795570399,
      "grad_norm": 0.3611898422241211,
      "learning_rate": 2e-05,
      "loss": 0.5353,
      "step": 6600
    },
    {
      "epoch": 0.5809456846545966,
      "grad_norm": 0.3862975239753723,
      "learning_rate": 2e-05,
      "loss": 0.5149,
      "step": 6610
    },
    {
      "epoch": 0.5818245737387941,
      "grad_norm": 0.35103002190589905,
      "learning_rate": 2e-05,
      "loss": 0.5041,
      "step": 6620
    },
    {
      "epoch": 0.5827034628229918,
      "grad_norm": 0.31625738739967346,
      "learning_rate": 2e-05,
      "loss": 0.522,
      "step": 6630
    },
    {
      "epoch": 0.5835823519071893,
      "grad_norm": 0.3532463014125824,
      "learning_rate": 2e-05,
      "loss": 0.5093,
      "step": 6640
    },
    {
      "epoch": 0.5844612409913869,
      "grad_norm": 0.3849062919616699,
      "learning_rate": 2e-05,
      "loss": 0.5269,
      "step": 6650
    },
    {
      "epoch": 0.5853401300755845,
      "grad_norm": 0.38369449973106384,
      "learning_rate": 2e-05,
      "loss": 0.5401,
      "step": 6660
    },
    {
      "epoch": 0.586219019159782,
      "grad_norm": 0.3854820132255554,
      "learning_rate": 2e-05,
      "loss": 0.54,
      "step": 6670
    },
    {
      "epoch": 0.5870979082439796,
      "grad_norm": 0.424874484539032,
      "learning_rate": 2e-05,
      "loss": 0.5431,
      "step": 6680
    },
    {
      "epoch": 0.5879767973281772,
      "grad_norm": 0.40439704060554504,
      "learning_rate": 2e-05,
      "loss": 0.5366,
      "step": 6690
    },
    {
      "epoch": 0.5888556864123747,
      "grad_norm": 0.5523659586906433,
      "learning_rate": 2e-05,
      "loss": 0.5303,
      "step": 6700
    },
    {
      "epoch": 0.5897345754965724,
      "grad_norm": 0.3525029420852661,
      "learning_rate": 2e-05,
      "loss": 0.5353,
      "step": 6710
    },
    {
      "epoch": 0.5906134645807699,
      "grad_norm": 0.39345619082450867,
      "learning_rate": 2e-05,
      "loss": 0.5665,
      "step": 6720
    },
    {
      "epoch": 0.5914923536649674,
      "grad_norm": 0.36468809843063354,
      "learning_rate": 2e-05,
      "loss": 0.5431,
      "step": 6730
    },
    {
      "epoch": 0.5923712427491651,
      "grad_norm": 0.3419267535209656,
      "learning_rate": 2e-05,
      "loss": 0.5385,
      "step": 6740
    },
    {
      "epoch": 0.5932501318333626,
      "grad_norm": 0.3630816340446472,
      "learning_rate": 2e-05,
      "loss": 0.5361,
      "step": 6750
    },
    {
      "epoch": 0.5941290209175603,
      "grad_norm": 0.3730619251728058,
      "learning_rate": 2e-05,
      "loss": 0.5257,
      "step": 6760
    },
    {
      "epoch": 0.5950079100017578,
      "grad_norm": 0.34850138425827026,
      "learning_rate": 2e-05,
      "loss": 0.524,
      "step": 6770
    },
    {
      "epoch": 0.5958867990859553,
      "grad_norm": 0.41456130146980286,
      "learning_rate": 2e-05,
      "loss": 0.5147,
      "step": 6780
    },
    {
      "epoch": 0.596765688170153,
      "grad_norm": 0.3912688195705414,
      "learning_rate": 2e-05,
      "loss": 0.5468,
      "step": 6790
    },
    {
      "epoch": 0.5976445772543505,
      "grad_norm": 0.35805875062942505,
      "learning_rate": 2e-05,
      "loss": 0.5238,
      "step": 6800
    },
    {
      "epoch": 0.598523466338548,
      "grad_norm": 0.33181798458099365,
      "learning_rate": 2e-05,
      "loss": 0.5164,
      "step": 6810
    },
    {
      "epoch": 0.5994023554227457,
      "grad_norm": 0.38323405385017395,
      "learning_rate": 2e-05,
      "loss": 0.5426,
      "step": 6820
    },
    {
      "epoch": 0.6002812445069432,
      "grad_norm": 0.4760487973690033,
      "learning_rate": 2e-05,
      "loss": 0.5397,
      "step": 6830
    },
    {
      "epoch": 0.6011601335911408,
      "grad_norm": 0.4131280779838562,
      "learning_rate": 2e-05,
      "loss": 0.5256,
      "step": 6840
    },
    {
      "epoch": 0.6020390226753384,
      "grad_norm": 0.3687100112438202,
      "learning_rate": 2e-05,
      "loss": 0.5132,
      "step": 6850
    },
    {
      "epoch": 0.6029179117595359,
      "grad_norm": 1.2745188474655151,
      "learning_rate": 2e-05,
      "loss": 0.5451,
      "step": 6860
    },
    {
      "epoch": 0.6037968008437335,
      "grad_norm": 0.3076283037662506,
      "learning_rate": 2e-05,
      "loss": 0.5217,
      "step": 6870
    },
    {
      "epoch": 0.6046756899279311,
      "grad_norm": 0.3894563317298889,
      "learning_rate": 2e-05,
      "loss": 0.5578,
      "step": 6880
    },
    {
      "epoch": 0.6055545790121287,
      "grad_norm": 0.3635370433330536,
      "learning_rate": 2e-05,
      "loss": 0.5415,
      "step": 6890
    },
    {
      "epoch": 0.6064334680963263,
      "grad_norm": 0.34984761476516724,
      "learning_rate": 2e-05,
      "loss": 0.5157,
      "step": 6900
    },
    {
      "epoch": 0.6073123571805238,
      "grad_norm": 0.3511723577976227,
      "learning_rate": 2e-05,
      "loss": 0.5246,
      "step": 6910
    },
    {
      "epoch": 0.6081912462647214,
      "grad_norm": 0.34231036901474,
      "learning_rate": 2e-05,
      "loss": 0.51,
      "step": 6920
    },
    {
      "epoch": 0.609070135348919,
      "grad_norm": 0.34825029969215393,
      "learning_rate": 2e-05,
      "loss": 0.5157,
      "step": 6930
    },
    {
      "epoch": 0.6099490244331165,
      "grad_norm": 0.37518739700317383,
      "learning_rate": 2e-05,
      "loss": 0.5184,
      "step": 6940
    },
    {
      "epoch": 0.6108279135173141,
      "grad_norm": 0.3367300033569336,
      "learning_rate": 2e-05,
      "loss": 0.5068,
      "step": 6950
    },
    {
      "epoch": 0.6117068026015117,
      "grad_norm": 0.32129815220832825,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 6960
    },
    {
      "epoch": 0.6125856916857093,
      "grad_norm": 0.3747527301311493,
      "learning_rate": 2e-05,
      "loss": 0.5146,
      "step": 6970
    },
    {
      "epoch": 0.6134645807699068,
      "grad_norm": 0.3887206017971039,
      "learning_rate": 2e-05,
      "loss": 0.5255,
      "step": 6980
    },
    {
      "epoch": 0.6143434698541044,
      "grad_norm": 0.4023424983024597,
      "learning_rate": 2e-05,
      "loss": 0.5258,
      "step": 6990
    },
    {
      "epoch": 0.615222358938302,
      "grad_norm": 0.4055207669734955,
      "learning_rate": 2e-05,
      "loss": 0.5313,
      "step": 7000
    },
    {
      "epoch": 0.6161012480224995,
      "grad_norm": 0.35139816999435425,
      "learning_rate": 2e-05,
      "loss": 0.5082,
      "step": 7010
    },
    {
      "epoch": 0.6169801371066971,
      "grad_norm": 0.34399065375328064,
      "learning_rate": 2e-05,
      "loss": 0.5174,
      "step": 7020
    },
    {
      "epoch": 0.6178590261908947,
      "grad_norm": 0.3247215449810028,
      "learning_rate": 2e-05,
      "loss": 0.5283,
      "step": 7030
    },
    {
      "epoch": 0.6187379152750923,
      "grad_norm": 0.36031219363212585,
      "learning_rate": 2e-05,
      "loss": 0.5266,
      "step": 7040
    },
    {
      "epoch": 0.6196168043592899,
      "grad_norm": 0.33793655037879944,
      "learning_rate": 2e-05,
      "loss": 0.5353,
      "step": 7050
    },
    {
      "epoch": 0.6204956934434874,
      "grad_norm": 0.4320240914821625,
      "learning_rate": 2e-05,
      "loss": 0.5246,
      "step": 7060
    },
    {
      "epoch": 0.621374582527685,
      "grad_norm": 0.33091577887535095,
      "learning_rate": 2e-05,
      "loss": 0.5215,
      "step": 7070
    },
    {
      "epoch": 0.6222534716118826,
      "grad_norm": 0.359786719083786,
      "learning_rate": 2e-05,
      "loss": 0.5261,
      "step": 7080
    },
    {
      "epoch": 0.6231323606960801,
      "grad_norm": 0.40029919147491455,
      "learning_rate": 2e-05,
      "loss": 0.5311,
      "step": 7090
    },
    {
      "epoch": 0.6240112497802778,
      "grad_norm": 0.3988870680332184,
      "learning_rate": 2e-05,
      "loss": 0.5355,
      "step": 7100
    },
    {
      "epoch": 0.6248901388644753,
      "grad_norm": 0.39657360315322876,
      "learning_rate": 2e-05,
      "loss": 0.535,
      "step": 7110
    },
    {
      "epoch": 0.6257690279486728,
      "grad_norm": 0.34774717688560486,
      "learning_rate": 2e-05,
      "loss": 0.545,
      "step": 7120
    },
    {
      "epoch": 0.6266479170328705,
      "grad_norm": 0.3459916114807129,
      "learning_rate": 2e-05,
      "loss": 0.5111,
      "step": 7130
    },
    {
      "epoch": 0.627526806117068,
      "grad_norm": 0.3749159574508667,
      "learning_rate": 2e-05,
      "loss": 0.5088,
      "step": 7140
    },
    {
      "epoch": 0.6284056952012655,
      "grad_norm": 0.3616851270198822,
      "learning_rate": 2e-05,
      "loss": 0.5041,
      "step": 7150
    },
    {
      "epoch": 0.6292845842854632,
      "grad_norm": 0.3790813386440277,
      "learning_rate": 2e-05,
      "loss": 0.5242,
      "step": 7160
    },
    {
      "epoch": 0.6301634733696607,
      "grad_norm": 0.37006130814552307,
      "learning_rate": 2e-05,
      "loss": 0.5233,
      "step": 7170
    },
    {
      "epoch": 0.6310423624538584,
      "grad_norm": 0.3337913155555725,
      "learning_rate": 2e-05,
      "loss": 0.5006,
      "step": 7180
    },
    {
      "epoch": 0.6319212515380559,
      "grad_norm": 0.37087202072143555,
      "learning_rate": 2e-05,
      "loss": 0.5288,
      "step": 7190
    },
    {
      "epoch": 0.6328001406222534,
      "grad_norm": 0.4053291082382202,
      "learning_rate": 2e-05,
      "loss": 0.511,
      "step": 7200
    },
    {
      "epoch": 0.6336790297064511,
      "grad_norm": 0.36167111992836,
      "learning_rate": 2e-05,
      "loss": 0.5268,
      "step": 7210
    },
    {
      "epoch": 0.6345579187906486,
      "grad_norm": 0.34520459175109863,
      "learning_rate": 2e-05,
      "loss": 0.5313,
      "step": 7220
    },
    {
      "epoch": 0.6354368078748462,
      "grad_norm": 0.41531798243522644,
      "learning_rate": 2e-05,
      "loss": 0.5285,
      "step": 7230
    },
    {
      "epoch": 0.6363156969590438,
      "grad_norm": 0.4149952828884125,
      "learning_rate": 2e-05,
      "loss": 0.5479,
      "step": 7240
    },
    {
      "epoch": 0.6371945860432413,
      "grad_norm": 0.3614427447319031,
      "learning_rate": 2e-05,
      "loss": 0.5278,
      "step": 7250
    },
    {
      "epoch": 0.638073475127439,
      "grad_norm": 0.39528965950012207,
      "learning_rate": 2e-05,
      "loss": 0.5196,
      "step": 7260
    },
    {
      "epoch": 0.6389523642116365,
      "grad_norm": 0.4004584550857544,
      "learning_rate": 2e-05,
      "loss": 0.5179,
      "step": 7270
    },
    {
      "epoch": 0.639831253295834,
      "grad_norm": 0.3738349974155426,
      "learning_rate": 2e-05,
      "loss": 0.5391,
      "step": 7280
    },
    {
      "epoch": 0.6407101423800317,
      "grad_norm": 0.37643641233444214,
      "learning_rate": 2e-05,
      "loss": 0.5459,
      "step": 7290
    },
    {
      "epoch": 0.6415890314642292,
      "grad_norm": 0.3787165582180023,
      "learning_rate": 2e-05,
      "loss": 0.5173,
      "step": 7300
    },
    {
      "epoch": 0.6424679205484268,
      "grad_norm": 0.3804172873497009,
      "learning_rate": 2e-05,
      "loss": 0.5589,
      "step": 7310
    },
    {
      "epoch": 0.6433468096326244,
      "grad_norm": 0.3884279727935791,
      "learning_rate": 2e-05,
      "loss": 0.5445,
      "step": 7320
    },
    {
      "epoch": 0.6442256987168219,
      "grad_norm": 0.3518822193145752,
      "learning_rate": 2e-05,
      "loss": 0.549,
      "step": 7330
    },
    {
      "epoch": 0.6451045878010195,
      "grad_norm": 0.3280639350414276,
      "learning_rate": 2e-05,
      "loss": 0.5246,
      "step": 7340
    },
    {
      "epoch": 0.6459834768852171,
      "grad_norm": 0.40235796570777893,
      "learning_rate": 2e-05,
      "loss": 0.545,
      "step": 7350
    },
    {
      "epoch": 0.6468623659694147,
      "grad_norm": 0.35578644275665283,
      "learning_rate": 2e-05,
      "loss": 0.5245,
      "step": 7360
    },
    {
      "epoch": 0.6477412550536122,
      "grad_norm": 0.3419851064682007,
      "learning_rate": 2e-05,
      "loss": 0.5026,
      "step": 7370
    },
    {
      "epoch": 0.6486201441378098,
      "grad_norm": 0.34933072328567505,
      "learning_rate": 2e-05,
      "loss": 0.5391,
      "step": 7380
    },
    {
      "epoch": 0.6494990332220074,
      "grad_norm": 0.67307448387146,
      "learning_rate": 2e-05,
      "loss": 0.5224,
      "step": 7390
    },
    {
      "epoch": 0.650377922306205,
      "grad_norm": 0.3876035809516907,
      "learning_rate": 2e-05,
      "loss": 0.531,
      "step": 7400
    },
    {
      "epoch": 0.6512568113904025,
      "grad_norm": 0.37991857528686523,
      "learning_rate": 2e-05,
      "loss": 0.5246,
      "step": 7410
    },
    {
      "epoch": 0.6521357004746001,
      "grad_norm": 0.34345942735671997,
      "learning_rate": 2e-05,
      "loss": 0.5481,
      "step": 7420
    },
    {
      "epoch": 0.6530145895587977,
      "grad_norm": 0.3234832286834717,
      "learning_rate": 2e-05,
      "loss": 0.5178,
      "step": 7430
    },
    {
      "epoch": 0.6538934786429953,
      "grad_norm": 0.37809255719184875,
      "learning_rate": 2e-05,
      "loss": 0.5256,
      "step": 7440
    },
    {
      "epoch": 0.6547723677271928,
      "grad_norm": 0.3648509979248047,
      "learning_rate": 2e-05,
      "loss": 0.5258,
      "step": 7450
    },
    {
      "epoch": 0.6556512568113904,
      "grad_norm": 0.3396671414375305,
      "learning_rate": 2e-05,
      "loss": 0.5027,
      "step": 7460
    },
    {
      "epoch": 0.656530145895588,
      "grad_norm": 0.35954615473747253,
      "learning_rate": 2e-05,
      "loss": 0.5268,
      "step": 7470
    },
    {
      "epoch": 0.6574090349797855,
      "grad_norm": 0.37251320481300354,
      "learning_rate": 2e-05,
      "loss": 0.5268,
      "step": 7480
    },
    {
      "epoch": 0.6582879240639832,
      "grad_norm": 0.3983251452445984,
      "learning_rate": 2e-05,
      "loss": 0.5295,
      "step": 7490
    },
    {
      "epoch": 0.6591668131481807,
      "grad_norm": 0.38986736536026,
      "learning_rate": 2e-05,
      "loss": 0.5128,
      "step": 7500
    },
    {
      "epoch": 0.6600457022323782,
      "grad_norm": 0.37529098987579346,
      "learning_rate": 2e-05,
      "loss": 0.5323,
      "step": 7510
    },
    {
      "epoch": 0.6609245913165759,
      "grad_norm": 0.3617500066757202,
      "learning_rate": 2e-05,
      "loss": 0.5365,
      "step": 7520
    },
    {
      "epoch": 0.6618034804007734,
      "grad_norm": 0.3870958685874939,
      "learning_rate": 2e-05,
      "loss": 0.5343,
      "step": 7530
    },
    {
      "epoch": 0.662682369484971,
      "grad_norm": 0.3328729569911957,
      "learning_rate": 2e-05,
      "loss": 0.5204,
      "step": 7540
    },
    {
      "epoch": 0.6635612585691686,
      "grad_norm": 0.3630615472793579,
      "learning_rate": 2e-05,
      "loss": 0.5378,
      "step": 7550
    },
    {
      "epoch": 0.6644401476533661,
      "grad_norm": 0.4269648790359497,
      "learning_rate": 2e-05,
      "loss": 0.539,
      "step": 7560
    },
    {
      "epoch": 0.6653190367375638,
      "grad_norm": 0.32995325326919556,
      "learning_rate": 2e-05,
      "loss": 0.5153,
      "step": 7570
    },
    {
      "epoch": 0.6661979258217613,
      "grad_norm": 0.38670384883880615,
      "learning_rate": 2e-05,
      "loss": 0.5179,
      "step": 7580
    },
    {
      "epoch": 0.6670768149059588,
      "grad_norm": 0.35918352007865906,
      "learning_rate": 2e-05,
      "loss": 0.5088,
      "step": 7590
    },
    {
      "epoch": 0.6679557039901565,
      "grad_norm": 0.37594136595726013,
      "learning_rate": 2e-05,
      "loss": 0.5681,
      "step": 7600
    },
    {
      "epoch": 0.668834593074354,
      "grad_norm": 0.3553025722503662,
      "learning_rate": 2e-05,
      "loss": 0.5138,
      "step": 7610
    },
    {
      "epoch": 0.6697134821585516,
      "grad_norm": 0.36668524146080017,
      "learning_rate": 2e-05,
      "loss": 0.531,
      "step": 7620
    },
    {
      "epoch": 0.6705923712427492,
      "grad_norm": 0.38124337792396545,
      "learning_rate": 2e-05,
      "loss": 0.5187,
      "step": 7630
    },
    {
      "epoch": 0.6714712603269467,
      "grad_norm": 0.34201160073280334,
      "learning_rate": 2e-05,
      "loss": 0.5256,
      "step": 7640
    },
    {
      "epoch": 0.6723501494111443,
      "grad_norm": 0.3732454776763916,
      "learning_rate": 2e-05,
      "loss": 0.5063,
      "step": 7650
    },
    {
      "epoch": 0.6732290384953419,
      "grad_norm": 0.34086358547210693,
      "learning_rate": 2e-05,
      "loss": 0.531,
      "step": 7660
    },
    {
      "epoch": 0.6741079275795394,
      "grad_norm": 0.4156498908996582,
      "learning_rate": 2e-05,
      "loss": 0.5155,
      "step": 7670
    },
    {
      "epoch": 0.674986816663737,
      "grad_norm": 0.3136572241783142,
      "learning_rate": 2e-05,
      "loss": 0.5258,
      "step": 7680
    },
    {
      "epoch": 0.6758657057479346,
      "grad_norm": 0.3465648293495178,
      "learning_rate": 2e-05,
      "loss": 0.4979,
      "step": 7690
    },
    {
      "epoch": 0.6767445948321322,
      "grad_norm": 0.38058042526245117,
      "learning_rate": 2e-05,
      "loss": 0.5344,
      "step": 7700
    },
    {
      "epoch": 0.6776234839163298,
      "grad_norm": 0.35429060459136963,
      "learning_rate": 2e-05,
      "loss": 0.5143,
      "step": 7710
    },
    {
      "epoch": 0.6785023730005273,
      "grad_norm": 0.39913210272789,
      "learning_rate": 2e-05,
      "loss": 0.5144,
      "step": 7720
    },
    {
      "epoch": 0.6793812620847249,
      "grad_norm": 0.35528939962387085,
      "learning_rate": 2e-05,
      "loss": 0.5627,
      "step": 7730
    },
    {
      "epoch": 0.6802601511689225,
      "grad_norm": 0.3690216541290283,
      "learning_rate": 2e-05,
      "loss": 0.5045,
      "step": 7740
    },
    {
      "epoch": 0.6811390402531201,
      "grad_norm": 0.38929465413093567,
      "learning_rate": 2e-05,
      "loss": 0.5493,
      "step": 7750
    },
    {
      "epoch": 0.6820179293373176,
      "grad_norm": 0.3343387842178345,
      "learning_rate": 2e-05,
      "loss": 0.5331,
      "step": 7760
    },
    {
      "epoch": 0.6828968184215152,
      "grad_norm": 0.40514594316482544,
      "learning_rate": 2e-05,
      "loss": 0.5282,
      "step": 7770
    },
    {
      "epoch": 0.6837757075057128,
      "grad_norm": 0.4368135631084442,
      "learning_rate": 2e-05,
      "loss": 0.5204,
      "step": 7780
    },
    {
      "epoch": 0.6846545965899103,
      "grad_norm": 0.3838955760002136,
      "learning_rate": 2e-05,
      "loss": 0.5404,
      "step": 7790
    },
    {
      "epoch": 0.6855334856741079,
      "grad_norm": 0.34718751907348633,
      "learning_rate": 2e-05,
      "loss": 0.5155,
      "step": 7800
    },
    {
      "epoch": 0.6864123747583055,
      "grad_norm": 0.3726848363876343,
      "learning_rate": 2e-05,
      "loss": 0.5086,
      "step": 7810
    },
    {
      "epoch": 0.687291263842503,
      "grad_norm": 0.32974350452423096,
      "learning_rate": 2e-05,
      "loss": 0.5181,
      "step": 7820
    },
    {
      "epoch": 0.6881701529267007,
      "grad_norm": 0.400511234998703,
      "learning_rate": 2e-05,
      "loss": 0.5073,
      "step": 7830
    },
    {
      "epoch": 0.6890490420108982,
      "grad_norm": 0.3527922034263611,
      "learning_rate": 2e-05,
      "loss": 0.515,
      "step": 7840
    },
    {
      "epoch": 0.6899279310950958,
      "grad_norm": 0.3599557876586914,
      "learning_rate": 2e-05,
      "loss": 0.5169,
      "step": 7850
    },
    {
      "epoch": 0.6908068201792934,
      "grad_norm": 0.37022262811660767,
      "learning_rate": 2e-05,
      "loss": 0.5029,
      "step": 7860
    },
    {
      "epoch": 0.6916857092634909,
      "grad_norm": 0.32002681493759155,
      "learning_rate": 2e-05,
      "loss": 0.5066,
      "step": 7870
    },
    {
      "epoch": 0.6925645983476886,
      "grad_norm": 0.34601953625679016,
      "learning_rate": 2e-05,
      "loss": 0.5214,
      "step": 7880
    },
    {
      "epoch": 0.6934434874318861,
      "grad_norm": 0.3614468276500702,
      "learning_rate": 2e-05,
      "loss": 0.5164,
      "step": 7890
    },
    {
      "epoch": 0.6943223765160836,
      "grad_norm": 0.34482458233833313,
      "learning_rate": 2e-05,
      "loss": 0.5096,
      "step": 7900
    },
    {
      "epoch": 0.6952012656002813,
      "grad_norm": 0.3786649703979492,
      "learning_rate": 2e-05,
      "loss": 0.5179,
      "step": 7910
    },
    {
      "epoch": 0.6960801546844788,
      "grad_norm": 0.394819974899292,
      "learning_rate": 2e-05,
      "loss": 0.5088,
      "step": 7920
    },
    {
      "epoch": 0.6969590437686763,
      "grad_norm": 0.35426345467567444,
      "learning_rate": 2e-05,
      "loss": 0.557,
      "step": 7930
    },
    {
      "epoch": 0.697837932852874,
      "grad_norm": 0.39353442192077637,
      "learning_rate": 2e-05,
      "loss": 0.5153,
      "step": 7940
    },
    {
      "epoch": 0.6987168219370715,
      "grad_norm": 0.3981497585773468,
      "learning_rate": 2e-05,
      "loss": 0.5585,
      "step": 7950
    },
    {
      "epoch": 0.6995957110212692,
      "grad_norm": 0.3652624785900116,
      "learning_rate": 2e-05,
      "loss": 0.5233,
      "step": 7960
    },
    {
      "epoch": 0.7004746001054667,
      "grad_norm": 0.40660157799720764,
      "learning_rate": 2e-05,
      "loss": 0.5464,
      "step": 7970
    },
    {
      "epoch": 0.7013534891896642,
      "grad_norm": 0.3289089500904083,
      "learning_rate": 2e-05,
      "loss": 0.4966,
      "step": 7980
    },
    {
      "epoch": 0.7022323782738619,
      "grad_norm": 0.3683902323246002,
      "learning_rate": 2e-05,
      "loss": 0.5107,
      "step": 7990
    },
    {
      "epoch": 0.7031112673580594,
      "grad_norm": 0.3746993839740753,
      "learning_rate": 2e-05,
      "loss": 0.5085,
      "step": 8000
    },
    {
      "epoch": 0.703990156442257,
      "grad_norm": 0.3916963040828705,
      "learning_rate": 2e-05,
      "loss": 0.5199,
      "step": 8010
    },
    {
      "epoch": 0.7048690455264546,
      "grad_norm": 0.3549078106880188,
      "learning_rate": 2e-05,
      "loss": 0.527,
      "step": 8020
    },
    {
      "epoch": 0.7057479346106521,
      "grad_norm": 0.4133472144603729,
      "learning_rate": 2e-05,
      "loss": 0.5105,
      "step": 8030
    },
    {
      "epoch": 0.7066268236948497,
      "grad_norm": 0.31728658080101013,
      "learning_rate": 2e-05,
      "loss": 0.5109,
      "step": 8040
    },
    {
      "epoch": 0.7075057127790473,
      "grad_norm": 0.36913833022117615,
      "learning_rate": 2e-05,
      "loss": 0.5085,
      "step": 8050
    },
    {
      "epoch": 0.7083846018632448,
      "grad_norm": 0.346512109041214,
      "learning_rate": 2e-05,
      "loss": 0.5055,
      "step": 8060
    },
    {
      "epoch": 0.7092634909474425,
      "grad_norm": 0.3483312427997589,
      "learning_rate": 2e-05,
      "loss": 0.5181,
      "step": 8070
    },
    {
      "epoch": 0.71014238003164,
      "grad_norm": 0.33441105484962463,
      "learning_rate": 2e-05,
      "loss": 0.5424,
      "step": 8080
    },
    {
      "epoch": 0.7110212691158376,
      "grad_norm": 0.39529186487197876,
      "learning_rate": 2e-05,
      "loss": 0.5198,
      "step": 8090
    },
    {
      "epoch": 0.7119001582000352,
      "grad_norm": 0.4085478186607361,
      "learning_rate": 2e-05,
      "loss": 0.5127,
      "step": 8100
    },
    {
      "epoch": 0.7127790472842327,
      "grad_norm": 0.3849215507507324,
      "learning_rate": 2e-05,
      "loss": 0.5238,
      "step": 8110
    },
    {
      "epoch": 0.7136579363684303,
      "grad_norm": 0.3753392994403839,
      "learning_rate": 2e-05,
      "loss": 0.5513,
      "step": 8120
    },
    {
      "epoch": 0.7145368254526279,
      "grad_norm": 0.33858412504196167,
      "learning_rate": 2e-05,
      "loss": 0.5047,
      "step": 8130
    },
    {
      "epoch": 0.7154157145368254,
      "grad_norm": 0.4012879431247711,
      "learning_rate": 2e-05,
      "loss": 0.5209,
      "step": 8140
    },
    {
      "epoch": 0.716294603621023,
      "grad_norm": 0.4335635006427765,
      "learning_rate": 2e-05,
      "loss": 0.5307,
      "step": 8150
    },
    {
      "epoch": 0.7171734927052206,
      "grad_norm": 0.38978126645088196,
      "learning_rate": 2e-05,
      "loss": 0.5285,
      "step": 8160
    },
    {
      "epoch": 0.7180523817894182,
      "grad_norm": 0.382772296667099,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 8170
    },
    {
      "epoch": 0.7189312708736157,
      "grad_norm": 0.3877055048942566,
      "learning_rate": 2e-05,
      "loss": 0.5184,
      "step": 8180
    },
    {
      "epoch": 0.7198101599578133,
      "grad_norm": 0.3682936728000641,
      "learning_rate": 2e-05,
      "loss": 0.4979,
      "step": 8190
    },
    {
      "epoch": 0.7206890490420109,
      "grad_norm": 0.3572366237640381,
      "learning_rate": 2e-05,
      "loss": 0.483,
      "step": 8200
    },
    {
      "epoch": 0.7215679381262085,
      "grad_norm": 0.37123164534568787,
      "learning_rate": 2e-05,
      "loss": 0.5146,
      "step": 8210
    },
    {
      "epoch": 0.7224468272104061,
      "grad_norm": 0.36580249667167664,
      "learning_rate": 2e-05,
      "loss": 0.529,
      "step": 8220
    },
    {
      "epoch": 0.7233257162946036,
      "grad_norm": 0.33150553703308105,
      "learning_rate": 2e-05,
      "loss": 0.4858,
      "step": 8230
    },
    {
      "epoch": 0.7242046053788012,
      "grad_norm": 0.3816601037979126,
      "learning_rate": 2e-05,
      "loss": 0.5322,
      "step": 8240
    },
    {
      "epoch": 0.7250834944629988,
      "grad_norm": 0.40648359060287476,
      "learning_rate": 2e-05,
      "loss": 0.5122,
      "step": 8250
    },
    {
      "epoch": 0.7259623835471963,
      "grad_norm": 0.3481503427028656,
      "learning_rate": 2e-05,
      "loss": 0.5079,
      "step": 8260
    },
    {
      "epoch": 0.7268412726313939,
      "grad_norm": 0.4123339056968689,
      "learning_rate": 2e-05,
      "loss": 0.5495,
      "step": 8270
    },
    {
      "epoch": 0.7277201617155915,
      "grad_norm": 0.34569188952445984,
      "learning_rate": 2e-05,
      "loss": 0.5363,
      "step": 8280
    },
    {
      "epoch": 0.728599050799789,
      "grad_norm": 0.33600306510925293,
      "learning_rate": 2e-05,
      "loss": 0.5086,
      "step": 8290
    },
    {
      "epoch": 0.7294779398839867,
      "grad_norm": 0.4178641736507416,
      "learning_rate": 2e-05,
      "loss": 0.5123,
      "step": 8300
    },
    {
      "epoch": 0.7303568289681842,
      "grad_norm": 0.34956926107406616,
      "learning_rate": 2e-05,
      "loss": 0.5143,
      "step": 8310
    },
    {
      "epoch": 0.7312357180523817,
      "grad_norm": 0.34772637486457825,
      "learning_rate": 2e-05,
      "loss": 0.5407,
      "step": 8320
    },
    {
      "epoch": 0.7321146071365794,
      "grad_norm": 0.36480712890625,
      "learning_rate": 2e-05,
      "loss": 0.5313,
      "step": 8330
    },
    {
      "epoch": 0.7329934962207769,
      "grad_norm": 0.3754159212112427,
      "learning_rate": 2e-05,
      "loss": 0.5046,
      "step": 8340
    },
    {
      "epoch": 0.7338723853049746,
      "grad_norm": 0.37473970651626587,
      "learning_rate": 2e-05,
      "loss": 0.5606,
      "step": 8350
    },
    {
      "epoch": 0.7347512743891721,
      "grad_norm": 0.3338927626609802,
      "learning_rate": 2e-05,
      "loss": 0.5281,
      "step": 8360
    },
    {
      "epoch": 0.7356301634733696,
      "grad_norm": 0.4149235486984253,
      "learning_rate": 2e-05,
      "loss": 0.5364,
      "step": 8370
    },
    {
      "epoch": 0.7365090525575673,
      "grad_norm": 0.3559480607509613,
      "learning_rate": 2e-05,
      "loss": 0.512,
      "step": 8380
    },
    {
      "epoch": 0.7373879416417648,
      "grad_norm": 0.34460002183914185,
      "learning_rate": 2e-05,
      "loss": 0.5125,
      "step": 8390
    },
    {
      "epoch": 0.7382668307259623,
      "grad_norm": 0.3763180375099182,
      "learning_rate": 2e-05,
      "loss": 0.5079,
      "step": 8400
    },
    {
      "epoch": 0.73914571981016,
      "grad_norm": 0.38870418071746826,
      "learning_rate": 2e-05,
      "loss": 0.513,
      "step": 8410
    },
    {
      "epoch": 0.7400246088943575,
      "grad_norm": 0.33556807041168213,
      "learning_rate": 2e-05,
      "loss": 0.506,
      "step": 8420
    },
    {
      "epoch": 0.7409034979785551,
      "grad_norm": 0.38033798336982727,
      "learning_rate": 2e-05,
      "loss": 0.499,
      "step": 8430
    },
    {
      "epoch": 0.7417823870627527,
      "grad_norm": 0.5256052017211914,
      "learning_rate": 2e-05,
      "loss": 0.528,
      "step": 8440
    },
    {
      "epoch": 0.7426612761469502,
      "grad_norm": 0.3809047043323517,
      "learning_rate": 2e-05,
      "loss": 0.4886,
      "step": 8450
    },
    {
      "epoch": 0.7435401652311479,
      "grad_norm": 0.38763684034347534,
      "learning_rate": 2e-05,
      "loss": 0.5164,
      "step": 8460
    },
    {
      "epoch": 0.7444190543153454,
      "grad_norm": 0.37231966853141785,
      "learning_rate": 2e-05,
      "loss": 0.5262,
      "step": 8470
    },
    {
      "epoch": 0.745297943399543,
      "grad_norm": 0.3402103781700134,
      "learning_rate": 2e-05,
      "loss": 0.5027,
      "step": 8480
    },
    {
      "epoch": 0.7461768324837406,
      "grad_norm": 0.38542816042900085,
      "learning_rate": 2e-05,
      "loss": 0.5215,
      "step": 8490
    },
    {
      "epoch": 0.7470557215679381,
      "grad_norm": 0.35029876232147217,
      "learning_rate": 2e-05,
      "loss": 0.5081,
      "step": 8500
    },
    {
      "epoch": 0.7479346106521357,
      "grad_norm": 0.40829455852508545,
      "learning_rate": 2e-05,
      "loss": 0.5554,
      "step": 8510
    },
    {
      "epoch": 0.7488134997363333,
      "grad_norm": 0.35434481501579285,
      "learning_rate": 2e-05,
      "loss": 0.5268,
      "step": 8520
    },
    {
      "epoch": 0.7496923888205308,
      "grad_norm": 0.36827224493026733,
      "learning_rate": 2e-05,
      "loss": 0.5409,
      "step": 8530
    },
    {
      "epoch": 0.7505712779047284,
      "grad_norm": 0.39406830072402954,
      "learning_rate": 2e-05,
      "loss": 0.5151,
      "step": 8540
    },
    {
      "epoch": 0.751450166988926,
      "grad_norm": 0.3718630373477936,
      "learning_rate": 2e-05,
      "loss": 0.5007,
      "step": 8550
    },
    {
      "epoch": 0.7523290560731236,
      "grad_norm": 0.3344067931175232,
      "learning_rate": 2e-05,
      "loss": 0.529,
      "step": 8560
    },
    {
      "epoch": 0.7532079451573211,
      "grad_norm": 0.3435877859592438,
      "learning_rate": 2e-05,
      "loss": 0.5037,
      "step": 8570
    },
    {
      "epoch": 0.7540868342415187,
      "grad_norm": 0.4093573987483978,
      "learning_rate": 2e-05,
      "loss": 0.5743,
      "step": 8580
    },
    {
      "epoch": 0.7549657233257163,
      "grad_norm": 0.36335745453834534,
      "learning_rate": 2e-05,
      "loss": 0.5381,
      "step": 8590
    },
    {
      "epoch": 0.7558446124099139,
      "grad_norm": 0.3628480136394501,
      "learning_rate": 2e-05,
      "loss": 0.5231,
      "step": 8600
    },
    {
      "epoch": 0.7567235014941115,
      "grad_norm": 0.3387520909309387,
      "learning_rate": 2e-05,
      "loss": 0.5311,
      "step": 8610
    },
    {
      "epoch": 0.757602390578309,
      "grad_norm": 0.3454357087612152,
      "learning_rate": 2e-05,
      "loss": 0.5164,
      "step": 8620
    },
    {
      "epoch": 0.7584812796625066,
      "grad_norm": 0.3372632563114166,
      "learning_rate": 2e-05,
      "loss": 0.5305,
      "step": 8630
    },
    {
      "epoch": 0.7593601687467042,
      "grad_norm": 0.3528139591217041,
      "learning_rate": 2e-05,
      "loss": 0.5213,
      "step": 8640
    },
    {
      "epoch": 0.7602390578309017,
      "grad_norm": 0.3492075502872467,
      "learning_rate": 2e-05,
      "loss": 0.5077,
      "step": 8650
    },
    {
      "epoch": 0.7611179469150993,
      "grad_norm": 0.3644988536834717,
      "learning_rate": 2e-05,
      "loss": 0.5105,
      "step": 8660
    },
    {
      "epoch": 0.7619968359992969,
      "grad_norm": 0.3513087034225464,
      "learning_rate": 2e-05,
      "loss": 0.4954,
      "step": 8670
    },
    {
      "epoch": 0.7628757250834944,
      "grad_norm": 0.3443455696105957,
      "learning_rate": 2e-05,
      "loss": 0.5131,
      "step": 8680
    },
    {
      "epoch": 0.7637546141676921,
      "grad_norm": 0.47363483905792236,
      "learning_rate": 2e-05,
      "loss": 0.5459,
      "step": 8690
    },
    {
      "epoch": 0.7646335032518896,
      "grad_norm": 0.335873007774353,
      "learning_rate": 2e-05,
      "loss": 0.5247,
      "step": 8700
    },
    {
      "epoch": 0.7655123923360871,
      "grad_norm": 0.3861873149871826,
      "learning_rate": 2e-05,
      "loss": 0.5208,
      "step": 8710
    },
    {
      "epoch": 0.7663912814202848,
      "grad_norm": 0.3416925072669983,
      "learning_rate": 2e-05,
      "loss": 0.5225,
      "step": 8720
    },
    {
      "epoch": 0.7672701705044823,
      "grad_norm": 0.3855714797973633,
      "learning_rate": 2e-05,
      "loss": 0.5248,
      "step": 8730
    },
    {
      "epoch": 0.76814905958868,
      "grad_norm": 0.36896610260009766,
      "learning_rate": 2e-05,
      "loss": 0.5089,
      "step": 8740
    },
    {
      "epoch": 0.7690279486728775,
      "grad_norm": 0.3854875862598419,
      "learning_rate": 2e-05,
      "loss": 0.5145,
      "step": 8750
    },
    {
      "epoch": 0.769906837757075,
      "grad_norm": 0.36021777987480164,
      "learning_rate": 2e-05,
      "loss": 0.5184,
      "step": 8760
    },
    {
      "epoch": 0.7707857268412727,
      "grad_norm": 0.32360488176345825,
      "learning_rate": 2e-05,
      "loss": 0.5175,
      "step": 8770
    },
    {
      "epoch": 0.7716646159254702,
      "grad_norm": 0.3740846812725067,
      "learning_rate": 2e-05,
      "loss": 0.5846,
      "step": 8780
    },
    {
      "epoch": 0.7725435050096677,
      "grad_norm": 0.36203324794769287,
      "learning_rate": 2e-05,
      "loss": 0.4895,
      "step": 8790
    },
    {
      "epoch": 0.7734223940938654,
      "grad_norm": 0.33148282766342163,
      "learning_rate": 2e-05,
      "loss": 0.4919,
      "step": 8800
    },
    {
      "epoch": 0.7743012831780629,
      "grad_norm": 0.408326119184494,
      "learning_rate": 2e-05,
      "loss": 0.5299,
      "step": 8810
    },
    {
      "epoch": 0.7751801722622605,
      "grad_norm": 0.4428783059120178,
      "learning_rate": 2e-05,
      "loss": 0.5217,
      "step": 8820
    },
    {
      "epoch": 0.7760590613464581,
      "grad_norm": 0.3390721082687378,
      "learning_rate": 2e-05,
      "loss": 0.5111,
      "step": 8830
    },
    {
      "epoch": 0.7769379504306556,
      "grad_norm": 0.3733547329902649,
      "learning_rate": 2e-05,
      "loss": 0.5504,
      "step": 8840
    },
    {
      "epoch": 0.7778168395148533,
      "grad_norm": 0.42180243134498596,
      "learning_rate": 2e-05,
      "loss": 0.5032,
      "step": 8850
    },
    {
      "epoch": 0.7786957285990508,
      "grad_norm": 0.39686283469200134,
      "learning_rate": 2e-05,
      "loss": 0.5428,
      "step": 8860
    },
    {
      "epoch": 0.7795746176832484,
      "grad_norm": 0.38678956031799316,
      "learning_rate": 2e-05,
      "loss": 0.496,
      "step": 8870
    },
    {
      "epoch": 0.780453506767446,
      "grad_norm": 0.35394591093063354,
      "learning_rate": 2e-05,
      "loss": 0.5174,
      "step": 8880
    },
    {
      "epoch": 0.7813323958516435,
      "grad_norm": 0.4041210412979126,
      "learning_rate": 2e-05,
      "loss": 0.5231,
      "step": 8890
    },
    {
      "epoch": 0.7822112849358411,
      "grad_norm": 0.38741424679756165,
      "learning_rate": 2e-05,
      "loss": 0.5249,
      "step": 8900
    },
    {
      "epoch": 0.7830901740200387,
      "grad_norm": 0.3606299161911011,
      "learning_rate": 2e-05,
      "loss": 0.5088,
      "step": 8910
    },
    {
      "epoch": 0.7839690631042362,
      "grad_norm": 0.3614653944969177,
      "learning_rate": 2e-05,
      "loss": 0.5297,
      "step": 8920
    },
    {
      "epoch": 0.7848479521884338,
      "grad_norm": 0.3372066020965576,
      "learning_rate": 2e-05,
      "loss": 0.498,
      "step": 8930
    },
    {
      "epoch": 0.7857268412726314,
      "grad_norm": 0.36030372977256775,
      "learning_rate": 2e-05,
      "loss": 0.5136,
      "step": 8940
    },
    {
      "epoch": 0.786605730356829,
      "grad_norm": 0.3689762055873871,
      "learning_rate": 2e-05,
      "loss": 0.5259,
      "step": 8950
    },
    {
      "epoch": 0.7874846194410265,
      "grad_norm": 0.36388272047042847,
      "learning_rate": 2e-05,
      "loss": 0.5122,
      "step": 8960
    },
    {
      "epoch": 0.7883635085252241,
      "grad_norm": 0.3534102439880371,
      "learning_rate": 2e-05,
      "loss": 0.517,
      "step": 8970
    },
    {
      "epoch": 0.7892423976094217,
      "grad_norm": 0.35463225841522217,
      "learning_rate": 2e-05,
      "loss": 0.5264,
      "step": 8980
    },
    {
      "epoch": 0.7901212866936193,
      "grad_norm": 0.46401548385620117,
      "learning_rate": 2e-05,
      "loss": 0.5291,
      "step": 8990
    },
    {
      "epoch": 0.7910001757778169,
      "grad_norm": 0.38110214471817017,
      "learning_rate": 2e-05,
      "loss": 0.5084,
      "step": 9000
    },
    {
      "epoch": 0.7918790648620144,
      "grad_norm": 0.35902294516563416,
      "learning_rate": 2e-05,
      "loss": 0.5359,
      "step": 9010
    },
    {
      "epoch": 0.792757953946212,
      "grad_norm": 0.3632373511791229,
      "learning_rate": 2e-05,
      "loss": 0.5211,
      "step": 9020
    },
    {
      "epoch": 0.7936368430304096,
      "grad_norm": 0.36524227261543274,
      "learning_rate": 2e-05,
      "loss": 0.5159,
      "step": 9030
    },
    {
      "epoch": 0.7945157321146071,
      "grad_norm": 0.4221189618110657,
      "learning_rate": 2e-05,
      "loss": 0.4998,
      "step": 9040
    },
    {
      "epoch": 0.7953946211988047,
      "grad_norm": 0.36077895760536194,
      "learning_rate": 2e-05,
      "loss": 0.5195,
      "step": 9050
    },
    {
      "epoch": 0.7962735102830023,
      "grad_norm": 0.33385607600212097,
      "learning_rate": 2e-05,
      "loss": 0.5054,
      "step": 9060
    },
    {
      "epoch": 0.7971523993671998,
      "grad_norm": 0.3818615972995758,
      "learning_rate": 2e-05,
      "loss": 0.4908,
      "step": 9070
    },
    {
      "epoch": 0.7980312884513975,
      "grad_norm": 0.39926251769065857,
      "learning_rate": 2e-05,
      "loss": 0.5058,
      "step": 9080
    },
    {
      "epoch": 0.798910177535595,
      "grad_norm": 0.3625210225582123,
      "learning_rate": 2e-05,
      "loss": 0.5231,
      "step": 9090
    },
    {
      "epoch": 0.7997890666197925,
      "grad_norm": 0.4178714156150818,
      "learning_rate": 2e-05,
      "loss": 0.5071,
      "step": 9100
    },
    {
      "epoch": 0.8006679557039902,
      "grad_norm": 0.39969193935394287,
      "learning_rate": 2e-05,
      "loss": 0.497,
      "step": 9110
    },
    {
      "epoch": 0.8015468447881877,
      "grad_norm": 0.31311678886413574,
      "learning_rate": 2e-05,
      "loss": 0.4838,
      "step": 9120
    },
    {
      "epoch": 0.8024257338723852,
      "grad_norm": 0.41818615794181824,
      "learning_rate": 2e-05,
      "loss": 0.5087,
      "step": 9130
    },
    {
      "epoch": 0.8033046229565829,
      "grad_norm": 0.35494768619537354,
      "learning_rate": 2e-05,
      "loss": 0.5227,
      "step": 9140
    },
    {
      "epoch": 0.8041835120407804,
      "grad_norm": 0.36232253909111023,
      "learning_rate": 2e-05,
      "loss": 0.5241,
      "step": 9150
    },
    {
      "epoch": 0.8050624011249781,
      "grad_norm": 0.3955244719982147,
      "learning_rate": 2e-05,
      "loss": 0.5423,
      "step": 9160
    },
    {
      "epoch": 0.8059412902091756,
      "grad_norm": 0.3635399043560028,
      "learning_rate": 2e-05,
      "loss": 0.5227,
      "step": 9170
    },
    {
      "epoch": 0.8068201792933731,
      "grad_norm": 0.33711761236190796,
      "learning_rate": 2e-05,
      "loss": 0.5047,
      "step": 9180
    },
    {
      "epoch": 0.8076990683775708,
      "grad_norm": 0.3646397888660431,
      "learning_rate": 2e-05,
      "loss": 0.527,
      "step": 9190
    },
    {
      "epoch": 0.8085779574617683,
      "grad_norm": 0.3489598333835602,
      "learning_rate": 2e-05,
      "loss": 0.5123,
      "step": 9200
    },
    {
      "epoch": 0.809456846545966,
      "grad_norm": 0.760221540927887,
      "learning_rate": 2e-05,
      "loss": 0.513,
      "step": 9210
    },
    {
      "epoch": 0.8103357356301635,
      "grad_norm": 0.363540917634964,
      "learning_rate": 2e-05,
      "loss": 0.5082,
      "step": 9220
    },
    {
      "epoch": 0.811214624714361,
      "grad_norm": 0.43587878346443176,
      "learning_rate": 2e-05,
      "loss": 0.5454,
      "step": 9230
    },
    {
      "epoch": 0.8120935137985587,
      "grad_norm": 0.35944098234176636,
      "learning_rate": 2e-05,
      "loss": 0.5213,
      "step": 9240
    },
    {
      "epoch": 0.8129724028827562,
      "grad_norm": 0.383098304271698,
      "learning_rate": 2e-05,
      "loss": 0.5058,
      "step": 9250
    },
    {
      "epoch": 0.8138512919669537,
      "grad_norm": 0.38402512669563293,
      "learning_rate": 2e-05,
      "loss": 0.5082,
      "step": 9260
    },
    {
      "epoch": 0.8147301810511514,
      "grad_norm": 0.3775666654109955,
      "learning_rate": 2e-05,
      "loss": 0.5104,
      "step": 9270
    },
    {
      "epoch": 0.8156090701353489,
      "grad_norm": 0.39746716618537903,
      "learning_rate": 2e-05,
      "loss": 0.5429,
      "step": 9280
    },
    {
      "epoch": 0.8164879592195465,
      "grad_norm": 0.37091633677482605,
      "learning_rate": 2e-05,
      "loss": 0.5243,
      "step": 9290
    },
    {
      "epoch": 0.8173668483037441,
      "grad_norm": 0.37809163331985474,
      "learning_rate": 2e-05,
      "loss": 0.5213,
      "step": 9300
    },
    {
      "epoch": 0.8182457373879416,
      "grad_norm": 0.384899377822876,
      "learning_rate": 2e-05,
      "loss": 0.5082,
      "step": 9310
    },
    {
      "epoch": 0.8191246264721392,
      "grad_norm": 0.3442026674747467,
      "learning_rate": 2e-05,
      "loss": 0.5222,
      "step": 9320
    },
    {
      "epoch": 0.8200035155563368,
      "grad_norm": 0.3657488524913788,
      "learning_rate": 2e-05,
      "loss": 0.5156,
      "step": 9330
    },
    {
      "epoch": 0.8208824046405344,
      "grad_norm": 0.37653082609176636,
      "learning_rate": 2e-05,
      "loss": 0.5134,
      "step": 9340
    },
    {
      "epoch": 0.821761293724732,
      "grad_norm": 0.3474156856536865,
      "learning_rate": 2e-05,
      "loss": 0.4979,
      "step": 9350
    },
    {
      "epoch": 0.8226401828089295,
      "grad_norm": 0.3877592384815216,
      "learning_rate": 2e-05,
      "loss": 0.511,
      "step": 9360
    },
    {
      "epoch": 0.8235190718931271,
      "grad_norm": 0.36288386583328247,
      "learning_rate": 2e-05,
      "loss": 0.5137,
      "step": 9370
    },
    {
      "epoch": 0.8243979609773247,
      "grad_norm": 0.40850698947906494,
      "learning_rate": 2e-05,
      "loss": 0.5093,
      "step": 9380
    },
    {
      "epoch": 0.8252768500615222,
      "grad_norm": 0.43312057852745056,
      "learning_rate": 2e-05,
      "loss": 0.5093,
      "step": 9390
    },
    {
      "epoch": 0.8261557391457198,
      "grad_norm": 0.49418747425079346,
      "learning_rate": 2e-05,
      "loss": 0.5126,
      "step": 9400
    },
    {
      "epoch": 0.8270346282299174,
      "grad_norm": 0.4172934889793396,
      "learning_rate": 2e-05,
      "loss": 0.5116,
      "step": 9410
    },
    {
      "epoch": 0.827913517314115,
      "grad_norm": 0.45830798149108887,
      "learning_rate": 2e-05,
      "loss": 0.5321,
      "step": 9420
    },
    {
      "epoch": 0.8287924063983125,
      "grad_norm": 0.35958370566368103,
      "learning_rate": 2e-05,
      "loss": 0.5197,
      "step": 9430
    },
    {
      "epoch": 0.8296712954825101,
      "grad_norm": 0.3545091450214386,
      "learning_rate": 2e-05,
      "loss": 0.4981,
      "step": 9440
    },
    {
      "epoch": 0.8305501845667077,
      "grad_norm": 0.3619720935821533,
      "learning_rate": 2e-05,
      "loss": 0.4884,
      "step": 9450
    },
    {
      "epoch": 0.8314290736509052,
      "grad_norm": 0.4328927993774414,
      "learning_rate": 2e-05,
      "loss": 0.5041,
      "step": 9460
    },
    {
      "epoch": 0.8323079627351029,
      "grad_norm": 0.36476585268974304,
      "learning_rate": 2e-05,
      "loss": 0.5189,
      "step": 9470
    },
    {
      "epoch": 0.8331868518193004,
      "grad_norm": 0.4025578498840332,
      "learning_rate": 2e-05,
      "loss": 0.5071,
      "step": 9480
    },
    {
      "epoch": 0.8340657409034979,
      "grad_norm": 0.3524565100669861,
      "learning_rate": 2e-05,
      "loss": 0.5112,
      "step": 9490
    },
    {
      "epoch": 0.8349446299876956,
      "grad_norm": 0.36523404717445374,
      "learning_rate": 2e-05,
      "loss": 0.5234,
      "step": 9500
    },
    {
      "epoch": 0.8358235190718931,
      "grad_norm": 0.33104559779167175,
      "learning_rate": 2e-05,
      "loss": 0.5151,
      "step": 9510
    },
    {
      "epoch": 0.8367024081560906,
      "grad_norm": 0.3614131212234497,
      "learning_rate": 2e-05,
      "loss": 0.528,
      "step": 9520
    },
    {
      "epoch": 0.8375812972402883,
      "grad_norm": 0.3441893756389618,
      "learning_rate": 2e-05,
      "loss": 0.4873,
      "step": 9530
    },
    {
      "epoch": 0.8384601863244858,
      "grad_norm": 0.3735648989677429,
      "learning_rate": 2e-05,
      "loss": 0.5116,
      "step": 9540
    },
    {
      "epoch": 0.8393390754086835,
      "grad_norm": 0.36411044001579285,
      "learning_rate": 2e-05,
      "loss": 0.4823,
      "step": 9550
    },
    {
      "epoch": 0.840217964492881,
      "grad_norm": 0.3676678240299225,
      "learning_rate": 2e-05,
      "loss": 0.5149,
      "step": 9560
    },
    {
      "epoch": 0.8410968535770785,
      "grad_norm": 0.39249342679977417,
      "learning_rate": 2e-05,
      "loss": 0.5135,
      "step": 9570
    },
    {
      "epoch": 0.8419757426612762,
      "grad_norm": 0.3674519956111908,
      "learning_rate": 2e-05,
      "loss": 0.522,
      "step": 9580
    },
    {
      "epoch": 0.8428546317454737,
      "grad_norm": 0.36224469542503357,
      "learning_rate": 2e-05,
      "loss": 0.4945,
      "step": 9590
    },
    {
      "epoch": 0.8437335208296713,
      "grad_norm": 0.3783111274242401,
      "learning_rate": 2e-05,
      "loss": 0.5128,
      "step": 9600
    },
    {
      "epoch": 0.8446124099138689,
      "grad_norm": 0.34140780568122864,
      "learning_rate": 2e-05,
      "loss": 0.4996,
      "step": 9610
    },
    {
      "epoch": 0.8454912989980664,
      "grad_norm": 0.40477558970451355,
      "learning_rate": 2e-05,
      "loss": 0.5286,
      "step": 9620
    },
    {
      "epoch": 0.846370188082264,
      "grad_norm": 0.36008432507514954,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 9630
    },
    {
      "epoch": 0.8472490771664616,
      "grad_norm": 0.33754611015319824,
      "learning_rate": 2e-05,
      "loss": 0.528,
      "step": 9640
    },
    {
      "epoch": 0.8481279662506591,
      "grad_norm": 0.4001263976097107,
      "learning_rate": 2e-05,
      "loss": 0.5301,
      "step": 9650
    },
    {
      "epoch": 0.8490068553348568,
      "grad_norm": 0.3425581455230713,
      "learning_rate": 2e-05,
      "loss": 0.4992,
      "step": 9660
    },
    {
      "epoch": 0.8498857444190543,
      "grad_norm": 0.44264301657676697,
      "learning_rate": 2e-05,
      "loss": 0.5237,
      "step": 9670
    },
    {
      "epoch": 0.8507646335032519,
      "grad_norm": 0.359456866979599,
      "learning_rate": 2e-05,
      "loss": 0.5275,
      "step": 9680
    },
    {
      "epoch": 0.8516435225874495,
      "grad_norm": 0.36370909214019775,
      "learning_rate": 2e-05,
      "loss": 0.5379,
      "step": 9690
    },
    {
      "epoch": 0.852522411671647,
      "grad_norm": 0.352681040763855,
      "learning_rate": 2e-05,
      "loss": 0.5172,
      "step": 9700
    },
    {
      "epoch": 0.8534013007558446,
      "grad_norm": 0.4034733176231384,
      "learning_rate": 2e-05,
      "loss": 0.532,
      "step": 9710
    },
    {
      "epoch": 0.8542801898400422,
      "grad_norm": 0.3679851293563843,
      "learning_rate": 2e-05,
      "loss": 0.535,
      "step": 9720
    },
    {
      "epoch": 0.8551590789242398,
      "grad_norm": 0.39791661500930786,
      "learning_rate": 2e-05,
      "loss": 0.5359,
      "step": 9730
    },
    {
      "epoch": 0.8560379680084373,
      "grad_norm": 0.3947438597679138,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 9740
    },
    {
      "epoch": 0.8569168570926349,
      "grad_norm": 0.3695662021636963,
      "learning_rate": 2e-05,
      "loss": 0.5072,
      "step": 9750
    },
    {
      "epoch": 0.8577957461768325,
      "grad_norm": 0.3508788049221039,
      "learning_rate": 2e-05,
      "loss": 0.527,
      "step": 9760
    },
    {
      "epoch": 0.85867463526103,
      "grad_norm": 0.37944233417510986,
      "learning_rate": 2e-05,
      "loss": 0.5029,
      "step": 9770
    },
    {
      "epoch": 0.8595535243452276,
      "grad_norm": 0.3189653754234314,
      "learning_rate": 2e-05,
      "loss": 0.4656,
      "step": 9780
    },
    {
      "epoch": 0.8604324134294252,
      "grad_norm": 0.4324260354042053,
      "learning_rate": 2e-05,
      "loss": 0.5508,
      "step": 9790
    },
    {
      "epoch": 0.8613113025136228,
      "grad_norm": 0.4262057840824127,
      "learning_rate": 2e-05,
      "loss": 0.4999,
      "step": 9800
    },
    {
      "epoch": 0.8621901915978204,
      "grad_norm": 0.4112226665019989,
      "learning_rate": 2e-05,
      "loss": 0.5338,
      "step": 9810
    },
    {
      "epoch": 0.8630690806820179,
      "grad_norm": 0.4202406704425812,
      "learning_rate": 2e-05,
      "loss": 0.5051,
      "step": 9820
    },
    {
      "epoch": 0.8639479697662155,
      "grad_norm": 0.4058152139186859,
      "learning_rate": 2e-05,
      "loss": 0.5088,
      "step": 9830
    },
    {
      "epoch": 0.8648268588504131,
      "grad_norm": 0.366022527217865,
      "learning_rate": 2e-05,
      "loss": 0.4987,
      "step": 9840
    },
    {
      "epoch": 0.8657057479346106,
      "grad_norm": 0.45777130126953125,
      "learning_rate": 2e-05,
      "loss": 0.5214,
      "step": 9850
    },
    {
      "epoch": 0.8665846370188083,
      "grad_norm": 0.36027228832244873,
      "learning_rate": 2e-05,
      "loss": 0.4966,
      "step": 9860
    },
    {
      "epoch": 0.8674635261030058,
      "grad_norm": 0.32471296191215515,
      "learning_rate": 2e-05,
      "loss": 0.4948,
      "step": 9870
    },
    {
      "epoch": 0.8683424151872033,
      "grad_norm": 0.395317405462265,
      "learning_rate": 2e-05,
      "loss": 0.5056,
      "step": 9880
    },
    {
      "epoch": 0.869221304271401,
      "grad_norm": 0.3986475467681885,
      "learning_rate": 2e-05,
      "loss": 0.5097,
      "step": 9890
    },
    {
      "epoch": 0.8701001933555985,
      "grad_norm": 0.4100594222545624,
      "learning_rate": 2e-05,
      "loss": 0.5544,
      "step": 9900
    },
    {
      "epoch": 0.870979082439796,
      "grad_norm": 0.3657548129558563,
      "learning_rate": 2e-05,
      "loss": 0.4634,
      "step": 9910
    },
    {
      "epoch": 0.8718579715239937,
      "grad_norm": 0.376322865486145,
      "learning_rate": 2e-05,
      "loss": 0.5051,
      "step": 9920
    },
    {
      "epoch": 0.8727368606081912,
      "grad_norm": 0.4205462336540222,
      "learning_rate": 2e-05,
      "loss": 0.5239,
      "step": 9930
    },
    {
      "epoch": 0.8736157496923889,
      "grad_norm": 0.37560635805130005,
      "learning_rate": 2e-05,
      "loss": 0.5022,
      "step": 9940
    },
    {
      "epoch": 0.8744946387765864,
      "grad_norm": 0.3754473328590393,
      "learning_rate": 2e-05,
      "loss": 0.5128,
      "step": 9950
    },
    {
      "epoch": 0.8753735278607839,
      "grad_norm": 0.36242303252220154,
      "learning_rate": 2e-05,
      "loss": 0.4916,
      "step": 9960
    },
    {
      "epoch": 0.8762524169449816,
      "grad_norm": 0.41383790969848633,
      "learning_rate": 2e-05,
      "loss": 0.5231,
      "step": 9970
    },
    {
      "epoch": 0.8771313060291791,
      "grad_norm": 0.3501926362514496,
      "learning_rate": 2e-05,
      "loss": 0.5075,
      "step": 9980
    },
    {
      "epoch": 0.8780101951133767,
      "grad_norm": 0.3248148560523987,
      "learning_rate": 2e-05,
      "loss": 0.484,
      "step": 9990
    },
    {
      "epoch": 0.8788890841975743,
      "grad_norm": 0.3496324419975281,
      "learning_rate": 2e-05,
      "loss": 0.5093,
      "step": 10000
    },
    {
      "epoch": 0.8797679732817718,
      "grad_norm": 0.3719414472579956,
      "learning_rate": 2e-05,
      "loss": 0.5166,
      "step": 10010
    },
    {
      "epoch": 0.8806468623659695,
      "grad_norm": 0.4309203028678894,
      "learning_rate": 2e-05,
      "loss": 0.5408,
      "step": 10020
    },
    {
      "epoch": 0.881525751450167,
      "grad_norm": 0.3745439946651459,
      "learning_rate": 2e-05,
      "loss": 0.504,
      "step": 10030
    },
    {
      "epoch": 0.8824046405343645,
      "grad_norm": 0.3733232617378235,
      "learning_rate": 2e-05,
      "loss": 0.5152,
      "step": 10040
    },
    {
      "epoch": 0.8832835296185622,
      "grad_norm": 0.40560439229011536,
      "learning_rate": 2e-05,
      "loss": 0.4853,
      "step": 10050
    },
    {
      "epoch": 0.8841624187027597,
      "grad_norm": 0.3813885450363159,
      "learning_rate": 2e-05,
      "loss": 0.5045,
      "step": 10060
    },
    {
      "epoch": 0.8850413077869573,
      "grad_norm": 0.354848176240921,
      "learning_rate": 2e-05,
      "loss": 0.4997,
      "step": 10070
    },
    {
      "epoch": 0.8859201968711549,
      "grad_norm": 0.42518213391304016,
      "learning_rate": 2e-05,
      "loss": 0.4993,
      "step": 10080
    },
    {
      "epoch": 0.8867990859553524,
      "grad_norm": 0.41859170794487,
      "learning_rate": 2e-05,
      "loss": 0.5175,
      "step": 10090
    },
    {
      "epoch": 0.88767797503955,
      "grad_norm": 0.33884570002555847,
      "learning_rate": 2e-05,
      "loss": 0.506,
      "step": 10100
    },
    {
      "epoch": 0.8885568641237476,
      "grad_norm": 0.38686126470565796,
      "learning_rate": 2e-05,
      "loss": 0.51,
      "step": 10110
    },
    {
      "epoch": 0.8894357532079452,
      "grad_norm": 0.35234925150871277,
      "learning_rate": 2e-05,
      "loss": 0.4916,
      "step": 10120
    },
    {
      "epoch": 0.8903146422921427,
      "grad_norm": 0.38117244839668274,
      "learning_rate": 2e-05,
      "loss": 0.4828,
      "step": 10130
    },
    {
      "epoch": 0.8911935313763403,
      "grad_norm": 0.32832470536231995,
      "learning_rate": 2e-05,
      "loss": 0.4952,
      "step": 10140
    },
    {
      "epoch": 0.8920724204605379,
      "grad_norm": 0.3908103108406067,
      "learning_rate": 2e-05,
      "loss": 0.4956,
      "step": 10150
    },
    {
      "epoch": 0.8929513095447354,
      "grad_norm": 0.4111677408218384,
      "learning_rate": 2e-05,
      "loss": 0.5116,
      "step": 10160
    },
    {
      "epoch": 0.893830198628933,
      "grad_norm": 0.44471821188926697,
      "learning_rate": 2e-05,
      "loss": 0.5162,
      "step": 10170
    },
    {
      "epoch": 0.8947090877131306,
      "grad_norm": 0.36384859681129456,
      "learning_rate": 1.999957274187931e-05,
      "loss": 0.4812,
      "step": 10180
    },
    {
      "epoch": 0.8955879767973282,
      "grad_norm": 0.4385080337524414,
      "learning_rate": 1.9991978062682906e-05,
      "loss": 0.5401,
      "step": 10190
    },
    {
      "epoch": 0.8964668658815258,
      "grad_norm": 0.41910427808761597,
      "learning_rate": 1.997489709987378e-05,
      "loss": 0.4903,
      "step": 10200
    },
    {
      "epoch": 0.8973457549657233,
      "grad_norm": 0.4035734236240387,
      "learning_rate": 1.9948346151502135e-05,
      "loss": 0.4897,
      "step": 10210
    },
    {
      "epoch": 0.8982246440499209,
      "grad_norm": 0.4231525659561157,
      "learning_rate": 1.991235055154325e-05,
      "loss": 0.5025,
      "step": 10220
    },
    {
      "epoch": 0.8991035331341185,
      "grad_norm": 0.3686576783657074,
      "learning_rate": 1.9866944645724693e-05,
      "loss": 0.4948,
      "step": 10230
    },
    {
      "epoch": 0.899982422218316,
      "grad_norm": 0.4149637222290039,
      "learning_rate": 1.9812171758754842e-05,
      "loss": 0.4818,
      "step": 10240
    },
    {
      "epoch": 0.9008613113025136,
      "grad_norm": 0.41059258580207825,
      "learning_rate": 1.9748084152983987e-05,
      "loss": 0.5134,
      "step": 10250
    },
    {
      "epoch": 0.9017402003867112,
      "grad_norm": 0.3917215168476105,
      "learning_rate": 1.967474297853744e-05,
      "loss": 0.5011,
      "step": 10260
    },
    {
      "epoch": 0.9026190894709087,
      "grad_norm": 0.392131507396698,
      "learning_rate": 1.9592218214968256e-05,
      "loss": 0.5206,
      "step": 10270
    },
    {
      "epoch": 0.9034979785551064,
      "grad_norm": 0.37240123748779297,
      "learning_rate": 1.9500588604485213e-05,
      "loss": 0.494,
      "step": 10280
    },
    {
      "epoch": 0.9043768676393039,
      "grad_norm": 0.35975173115730286,
      "learning_rate": 1.939994157681979e-05,
      "loss": 0.4955,
      "step": 10290
    },
    {
      "epoch": 0.9052557567235014,
      "grad_norm": 0.40771472454071045,
      "learning_rate": 1.9290373165803803e-05,
      "loss": 0.4982,
      "step": 10300
    },
    {
      "epoch": 0.9061346458076991,
      "grad_norm": 1.3417563438415527,
      "learning_rate": 1.9171987917737328e-05,
      "loss": 0.5234,
      "step": 10310
    },
    {
      "epoch": 0.9070135348918966,
      "grad_norm": 0.40896889567375183,
      "learning_rate": 1.9044898791634322e-05,
      "loss": 0.4737,
      "step": 10320
    },
    {
      "epoch": 0.9078924239760943,
      "grad_norm": 0.5089636445045471,
      "learning_rate": 1.8909227051441126e-05,
      "loss": 0.5155,
      "step": 10330
    },
    {
      "epoch": 0.9087713130602918,
      "grad_norm": 0.47792020440101624,
      "learning_rate": 1.8765102150330692e-05,
      "loss": 0.5113,
      "step": 10340
    },
    {
      "epoch": 0.9096502021444893,
      "grad_norm": 0.3719557225704193,
      "learning_rate": 1.861266160718297e-05,
      "loss": 0.4818,
      "step": 10350
    },
    {
      "epoch": 0.910529091228687,
      "grad_norm": 0.3631979525089264,
      "learning_rate": 1.845205087536923e-05,
      "loss": 0.5131,
      "step": 10360
    },
    {
      "epoch": 0.9114079803128845,
      "grad_norm": 0.42810317873954773,
      "learning_rate": 1.8283423203965633e-05,
      "loss": 0.5263,
      "step": 10370
    },
    {
      "epoch": 0.912286869397082,
      "grad_norm": 0.36269620060920715,
      "learning_rate": 1.810693949152835e-05,
      "loss": 0.5096,
      "step": 10380
    },
    {
      "epoch": 0.9131657584812797,
      "grad_norm": 0.39026758074760437,
      "learning_rate": 1.7922768132569902e-05,
      "loss": 0.506,
      "step": 10390
    },
    {
      "epoch": 0.9140446475654772,
      "grad_norm": 0.3627564609050751,
      "learning_rate": 1.773108485688302e-05,
      "loss": 0.4881,
      "step": 10400
    },
    {
      "epoch": 0.9149235366496749,
      "grad_norm": 0.3748854398727417,
      "learning_rate": 1.7532072561865537e-05,
      "loss": 0.4813,
      "step": 10410
    },
    {
      "epoch": 0.9158024257338724,
      "grad_norm": 0.3798271119594574,
      "learning_rate": 1.732592113800616e-05,
      "loss": 0.5385,
      "step": 10420
    },
    {
      "epoch": 0.9166813148180699,
      "grad_norm": 0.3728078007698059,
      "learning_rate": 1.7112827287697668e-05,
      "loss": 0.5073,
      "step": 10430
    },
    {
      "epoch": 0.9175602039022676,
      "grad_norm": 0.35570967197418213,
      "learning_rate": 1.6892994337550452e-05,
      "loss": 0.496,
      "step": 10440
    },
    {
      "epoch": 0.9184390929864651,
      "grad_norm": 0.36647963523864746,
      "learning_rate": 1.6666632044385487e-05,
      "loss": 0.4619,
      "step": 10450
    },
    {
      "epoch": 0.9193179820706627,
      "grad_norm": 0.4051192104816437,
      "learning_rate": 1.6433956395091757e-05,
      "loss": 0.5101,
      "step": 10460
    },
    {
      "epoch": 0.9201968711548603,
      "grad_norm": 0.33733633160591125,
      "learning_rate": 1.6195189400539238e-05,
      "loss": 0.4688,
      "step": 10470
    },
    {
      "epoch": 0.9210757602390578,
      "grad_norm": 0.34595632553100586,
      "learning_rate": 1.595055888374398e-05,
      "loss": 0.4876,
      "step": 10480
    },
    {
      "epoch": 0.9219546493232554,
      "grad_norm": 0.4237731397151947,
      "learning_rate": 1.5700298262487434e-05,
      "loss": 0.4823,
      "step": 10490
    },
    {
      "epoch": 0.922833538407453,
      "grad_norm": 0.3246361017227173,
      "learning_rate": 1.5444646326597467e-05,
      "loss": 0.5025,
      "step": 10500
    },
    {
      "epoch": 0.9237124274916505,
      "grad_norm": 0.3979153633117676,
      "learning_rate": 1.5183847010103586e-05,
      "loss": 0.5145,
      "step": 10510
    },
    {
      "epoch": 0.9245913165758481,
      "grad_norm": 0.392475962638855,
      "learning_rate": 1.4918149158483686e-05,
      "loss": 0.5028,
      "step": 10520
    },
    {
      "epoch": 0.9254702056600457,
      "grad_norm": 0.3925435543060303,
      "learning_rate": 1.464780629122452e-05,
      "loss": 0.4992,
      "step": 10530
    },
    {
      "epoch": 0.9263490947442433,
      "grad_norm": 0.338585764169693,
      "learning_rate": 1.4373076359922374e-05,
      "loss": 0.4927,
      "step": 10540
    },
    {
      "epoch": 0.9272279838284408,
      "grad_norm": 0.3975823223590851,
      "learning_rate": 1.4094221502154774e-05,
      "loss": 0.5083,
      "step": 10550
    },
    {
      "epoch": 0.9281068729126384,
      "grad_norm": 0.3656439483165741,
      "learning_rate": 1.3811507791358113e-05,
      "loss": 0.4869,
      "step": 10560
    },
    {
      "epoch": 0.928985761996836,
      "grad_norm": 0.37011364102363586,
      "learning_rate": 1.3525204982949753e-05,
      "loss": 0.5013,
      "step": 10570
    },
    {
      "epoch": 0.9298646510810336,
      "grad_norm": 0.34798765182495117,
      "learning_rate": 1.3235586256936996e-05,
      "loss": 0.5006,
      "step": 10580
    },
    {
      "epoch": 0.9307435401652312,
      "grad_norm": 0.36794590950012207,
      "learning_rate": 1.294292795725836e-05,
      "loss": 0.4937,
      "step": 10590
    },
    {
      "epoch": 0.9316224292494287,
      "grad_norm": 0.35447412729263306,
      "learning_rate": 1.2647509328105978e-05,
      "loss": 0.5069,
      "step": 10600
    },
    {
      "epoch": 0.9325013183336263,
      "grad_norm": 0.4333643913269043,
      "learning_rate": 1.2349612247480686e-05,
      "loss": 0.4933,
      "step": 10610
    },
    {
      "epoch": 0.9333802074178239,
      "grad_norm": 0.3556787967681885,
      "learning_rate": 1.2049520958234e-05,
      "loss": 0.5122,
      "step": 10620
    },
    {
      "epoch": 0.9342590965020214,
      "grad_norm": 0.4080731272697449,
      "learning_rate": 1.1747521796853638e-05,
      "loss": 0.5019,
      "step": 10630
    },
    {
      "epoch": 0.935137985586219,
      "grad_norm": 0.3613244891166687,
      "learning_rate": 1.1443902920251395e-05,
      "loss": 0.503,
      "step": 10640
    },
    {
      "epoch": 0.9360168746704166,
      "grad_norm": 0.37831804156303406,
      "learning_rate": 1.1138954030814009e-05,
      "loss": 0.5145,
      "step": 10650
    },
    {
      "epoch": 0.9368957637546141,
      "grad_norm": 0.36623135209083557,
      "learning_rate": 1.0832966099979388e-05,
      "loss": 0.5265,
      "step": 10660
    },
    {
      "epoch": 0.9377746528388118,
      "grad_norm": 0.30365148186683655,
      "learning_rate": 1.0526231090601969e-05,
      "loss": 0.5029,
      "step": 10670
    },
    {
      "epoch": 0.9386535419230093,
      "grad_norm": 0.37851569056510925,
      "learning_rate": 1.0219041678372106e-05,
      "loss": 0.4974,
      "step": 10680
    },
    {
      "epoch": 0.9395324310072068,
      "grad_norm": 0.7586756348609924,
      "learning_rate": 9.911690972555268e-06,
      "loss": 0.4907,
      "step": 10690
    },
    {
      "epoch": 0.9404113200914045,
      "grad_norm": 0.37997883558273315,
      "learning_rate": 9.604472236317541e-06,
      "loss": 0.5035,
      "step": 10700
    },
    {
      "epoch": 0.941290209175602,
      "grad_norm": 0.3712542653083801,
      "learning_rate": 9.297678606904335e-06,
      "loss": 0.5111,
      "step": 10710
    },
    {
      "epoch": 0.9421690982597997,
      "grad_norm": 0.3893560469150543,
      "learning_rate": 8.991602815939155e-06,
      "loss": 0.5177,
      "step": 10720
    },
    {
      "epoch": 0.9430479873439972,
      "grad_norm": 0.39914897084236145,
      "learning_rate": 8.686536910109464e-06,
      "loss": 0.4843,
      "step": 10730
    },
    {
      "epoch": 0.9439268764281947,
      "grad_norm": 0.35491764545440674,
      "learning_rate": 8.382771972506084e-06,
      "loss": 0.5026,
      "step": 10740
    },
    {
      "epoch": 0.9448057655123924,
      "grad_norm": 0.34292975068092346,
      "learning_rate": 8.08059784488198e-06,
      "loss": 0.4933,
      "step": 10750
    },
    {
      "epoch": 0.9456846545965899,
      "grad_norm": 0.34541940689086914,
      "learning_rate": 7.780302851095544e-06,
      "loss": 0.4991,
      "step": 10760
    },
    {
      "epoch": 0.9465635436807874,
      "grad_norm": 0.38087719678878784,
      "learning_rate": 7.482173522002166e-06,
      "loss": 0.4951,
      "step": 10770
    },
    {
      "epoch": 0.9474424327649851,
      "grad_norm": 0.3610073924064636,
      "learning_rate": 7.186494322056624e-06,
      "loss": 0.5005,
      "step": 10780
    },
    {
      "epoch": 0.9483213218491826,
      "grad_norm": 0.3650810718536377,
      "learning_rate": 6.893547377887183e-06,
      "loss": 0.5222,
      "step": 10790
    },
    {
      "epoch": 0.9492002109333803,
      "grad_norm": 0.5252271294593811,
      "learning_rate": 6.6036122091003365e-06,
      "loss": 0.4918,
      "step": 10800
    },
    {
      "epoch": 0.9500791000175778,
      "grad_norm": 0.3804275095462799,
      "learning_rate": 6.316965461573097e-06,
      "loss": 0.4599,
      "step": 10810
    },
    {
      "epoch": 0.9509579891017753,
      "grad_norm": 0.3524964451789856,
      "learning_rate": 6.0338806434872965e-06,
      "loss": 0.4874,
      "step": 10820
    },
    {
      "epoch": 0.951836878185973,
      "grad_norm": 0.35691070556640625,
      "learning_rate": 5.754627864357749e-06,
      "loss": 0.4992,
      "step": 10830
    },
    {
      "epoch": 0.9527157672701705,
      "grad_norm": 0.3480913043022156,
      "learning_rate": 5.479473577303319e-06,
      "loss": 0.4975,
      "step": 10840
    },
    {
      "epoch": 0.9535946563543681,
      "grad_norm": 0.36395952105522156,
      "learning_rate": 5.208680324806778e-06,
      "loss": 0.4614,
      "step": 10850
    },
    {
      "epoch": 0.9544735454385657,
      "grad_norm": 0.41447848081588745,
      "learning_rate": 4.94250648820606e-06,
      "loss": 0.5088,
      "step": 10860
    },
    {
      "epoch": 0.9553524345227632,
      "grad_norm": 0.37823230028152466,
      "learning_rate": 4.681206041155951e-06,
      "loss": 0.5137,
      "step": 10870
    },
    {
      "epoch": 0.9562313236069608,
      "grad_norm": 0.38601452112197876,
      "learning_rate": 4.42502830729538e-06,
      "loss": 0.4756,
      "step": 10880
    },
    {
      "epoch": 0.9571102126911584,
      "grad_norm": 0.4184951186180115,
      "learning_rate": 4.174217722351669e-06,
      "loss": 0.4932,
      "step": 10890
    },
    {
      "epoch": 0.9579891017753559,
      "grad_norm": 0.4012276530265808,
      "learning_rate": 3.929013600908575e-06,
      "loss": 0.5493,
      "step": 10900
    },
    {
      "epoch": 0.9588679908595535,
      "grad_norm": 0.40644392371177673,
      "learning_rate": 3.689649908060807e-06,
      "loss": 0.4917,
      "step": 10910
    },
    {
      "epoch": 0.9597468799437511,
      "grad_norm": 0.39161422848701477,
      "learning_rate": 3.4563550361728028e-06,
      "loss": 0.486,
      "step": 10920
    },
    {
      "epoch": 0.9606257690279487,
      "grad_norm": 0.483493447303772,
      "learning_rate": 3.2293515869548216e-06,
      "loss": 0.5112,
      "step": 10930
    },
    {
      "epoch": 0.9615046581121462,
      "grad_norm": 0.39913806319236755,
      "learning_rate": 3.0088561590642746e-06,
      "loss": 0.5055,
      "step": 10940
    },
    {
      "epoch": 0.9623835471963438,
      "grad_norm": 0.3435784876346588,
      "learning_rate": 2.795079141434941e-06,
      "loss": 0.5218,
      "step": 10950
    },
    {
      "epoch": 0.9632624362805414,
      "grad_norm": 0.38981232047080994,
      "learning_rate": 2.588224512531299e-06,
      "loss": 0.4742,
      "step": 10960
    },
    {
      "epoch": 0.964141325364739,
      "grad_norm": 0.347115159034729,
      "learning_rate": 2.388489645719495e-06,
      "loss": 0.4886,
      "step": 10970
    },
    {
      "epoch": 0.9650202144489366,
      "grad_norm": 0.42196497321128845,
      "learning_rate": 2.196065120940663e-06,
      "loss": 0.5115,
      "step": 10980
    },
    {
      "epoch": 0.9658991035331341,
      "grad_norm": 0.3616836965084076,
      "learning_rate": 2.0111345428662808e-06,
      "loss": 0.4943,
      "step": 10990
    },
    {
      "epoch": 0.9667779926173317,
      "grad_norm": 0.3918362557888031,
      "learning_rate": 1.8338743657091102e-06,
      "loss": 0.4843,
      "step": 11000
    },
    {
      "epoch": 0.9676568817015293,
      "grad_norm": 0.3631676137447357,
      "learning_rate": 1.6644537248568113e-06,
      "loss": 0.4824,
      "step": 11010
    },
    {
      "epoch": 0.9685357707857268,
      "grad_norm": 0.3569515347480774,
      "learning_rate": 1.503034275488951e-06,
      "loss": 0.4943,
      "step": 11020
    },
    {
      "epoch": 0.9694146598699244,
      "grad_norm": 0.4705686867237091,
      "learning_rate": 1.3497700383313495e-06,
      "loss": 0.5191,
      "step": 11030
    },
    {
      "epoch": 0.970293548954122,
      "grad_norm": 0.34463125467300415,
      "learning_rate": 1.2048072526949562e-06,
      "loss": 0.4684,
      "step": 11040
    },
    {
      "epoch": 0.9711724380383195,
      "grad_norm": 0.3724505603313446,
      "learning_rate": 1.0682842369394756e-06,
      "loss": 0.5136,
      "step": 11050
    },
    {
      "epoch": 0.9720513271225172,
      "grad_norm": 0.2930002212524414,
      "learning_rate": 9.403312564948792e-07,
      "loss": 0.5116,
      "step": 11060
    },
    {
      "epoch": 0.9729302162067147,
      "grad_norm": 0.49282699823379517,
      "learning_rate": 8.210703995667471e-07,
      "loss": 0.5134,
      "step": 11070
    },
    {
      "epoch": 0.9738091052909122,
      "grad_norm": 0.35777708888053894,
      "learning_rate": 7.106154606440245e-07,
      "loss": 0.4727,
      "step": 11080
    },
    {
      "epoch": 0.9746879943751099,
      "grad_norm": 0.33871257305145264,
      "learning_rate": 6.090718319203393e-07,
      "loss": 0.4879,
      "step": 11090
    },
    {
      "epoch": 0.9755668834593074,
      "grad_norm": 0.40422382950782776,
      "learning_rate": 5.165364027325089e-07,
      "loss": 0.5024,
      "step": 11100
    },
    {
      "epoch": 0.9764457725435051,
      "grad_norm": 0.335046648979187,
      "learning_rate": 4.330974671121585e-07,
      "loss": 0.5253,
      "step": 11110
    },
    {
      "epoch": 0.9773246616277026,
      "grad_norm": 0.3367549479007721,
      "learning_rate": 3.588346395386846e-07,
      "loss": 0.491,
      "step": 11120
    },
    {
      "epoch": 0.9782035507119001,
      "grad_norm": 0.35378339886665344,
      "learning_rate": 2.938187789739395e-07,
      "loss": 0.4768,
      "step": 11130
    },
    {
      "epoch": 0.9790824397960978,
      "grad_norm": 0.386150062084198,
      "learning_rate": 2.3811192125110724e-07,
      "loss": 0.4836,
      "step": 11140
    },
    {
      "epoch": 0.9799613288802953,
      "grad_norm": 0.3469489812850952,
      "learning_rate": 1.917672198823177e-07,
      "loss": 0.4932,
      "step": 11150
    },
    {
      "epoch": 0.9808402179644928,
      "grad_norm": 0.3448003828525543,
      "learning_rate": 1.5482889534143795e-07,
      "loss": 0.4961,
      "step": 11160
    },
    {
      "epoch": 0.9817191070486905,
      "grad_norm": 0.3894102871417999,
      "learning_rate": 1.273321928704694e-07,
      "loss": 0.4933,
      "step": 11170
    },
    {
      "epoch": 0.982597996132888,
      "grad_norm": 0.36532095074653625,
      "learning_rate": 1.0930334884977866e-07,
      "loss": 0.497,
      "step": 11180
    },
    {
      "epoch": 0.9834768852170857,
      "grad_norm": 0.37706083059310913,
      "learning_rate": 1.0075956576427487e-07,
      "loss": 0.5312,
      "step": 11190
    },
    {
      "epoch": 0.9837405519423449,
      "step": 11193,
      "total_flos": 4.9965855215861105e+19,
      "train_loss": 0.04536843330751717,
      "train_runtime": 11372.8418,
      "train_samples_per_second": 11.81,
      "train_steps_per_second": 0.984
    }
  ],
  "logging_steps": 10,
  "max_steps": 11193,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 636,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.9965855215861105e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}