{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.24582527685006153,
  "eval_steps": 500,
  "global_step": 2797,
  "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": 0.3850197196006775,
      "learning_rate": 1.997261400182901e-05,
      "loss": 0.5326,
      "step": 2550
    },
    {
      "epoch": 0.224995605554579,
      "grad_norm": 0.3757331073284149,
      "learning_rate": 1.9805800746810423e-05,
      "loss": 0.5473,
      "step": 2560
    },
    {
      "epoch": 0.2258744946387766,
      "grad_norm": 0.37675556540489197,
      "learning_rate": 1.9489934220110274e-05,
      "loss": 0.5473,
      "step": 2570
    },
    {
      "epoch": 0.22675338372297416,
      "grad_norm": 0.33922237157821655,
      "learning_rate": 1.9029840364820392e-05,
      "loss": 0.5686,
      "step": 2580
    },
    {
      "epoch": 0.22763227280717174,
      "grad_norm": 0.4285179674625397,
      "learning_rate": 1.8432548690479908e-05,
      "loss": 0.5627,
      "step": 2590
    },
    {
      "epoch": 0.2285111618913693,
      "grad_norm": 0.3418651521205902,
      "learning_rate": 1.7707184873239772e-05,
      "loss": 0.5532,
      "step": 2600
    },
    {
      "epoch": 0.2293900509755669,
      "grad_norm": 0.32062989473342896,
      "learning_rate": 1.6864831329902565e-05,
      "loss": 0.5297,
      "step": 2610
    },
    {
      "epoch": 0.23026894005976445,
      "grad_norm": 0.36912357807159424,
      "learning_rate": 1.591835789604667e-05,
      "loss": 0.5711,
      "step": 2620
    },
    {
      "epoch": 0.23114782914396204,
      "grad_norm": 0.3297048807144165,
      "learning_rate": 1.4882225195206459e-05,
      "loss": 0.5534,
      "step": 2630
    },
    {
      "epoch": 0.2320267182281596,
      "grad_norm": 0.3525194823741913,
      "learning_rate": 1.3772263703317695e-05,
      "loss": 0.58,
      "step": 2640
    },
    {
      "epoch": 0.23290560731235718,
      "grad_norm": 0.34920766949653625,
      "learning_rate": 1.2605431883975547e-05,
      "loss": 0.5545,
      "step": 2650
    },
    {
      "epoch": 0.23378449639655474,
      "grad_norm": 0.34322798252105713,
      "learning_rate": 1.1399557089817574e-05,
      "loss": 0.5629,
      "step": 2660
    },
    {
      "epoch": 0.23466338548075233,
      "grad_norm": 0.354938805103302,
      "learning_rate": 1.0173063188650463e-05,
      "loss": 0.5595,
      "step": 2670
    },
    {
      "epoch": 0.23554227456494992,
      "grad_norm": 0.32228320837020874,
      "learning_rate": 8.944689075764396e-06,
      "loss": 0.5707,
      "step": 2680
    },
    {
      "epoch": 0.23642116364914748,
      "grad_norm": 0.34030893445014954,
      "learning_rate": 7.73320237312362e-06,
      "loss": 0.5602,
      "step": 2690
    },
    {
      "epoch": 0.23730005273334506,
      "grad_norm": 0.3501499891281128,
      "learning_rate": 6.557112689658629e-06,
      "loss": 0.5712,
      "step": 2700
    },
    {
      "epoch": 0.23817894181754262,
      "grad_norm": 0.3431225121021271,
      "learning_rate": 5.434388823591213e-06,
      "loss": 0.5775,
      "step": 2710
    },
    {
      "epoch": 0.2390578309017402,
      "grad_norm": 0.34175869822502136,
      "learning_rate": 4.382184227495424e-06,
      "loss": 0.5441,
      "step": 2720
    },
    {
      "epoch": 0.23993671998593777,
      "grad_norm": 0.37560412287712097,
      "learning_rate": 3.4165749305560203e-06,
      "loss": 0.5298,
      "step": 2730
    },
    {
      "epoch": 0.24081560907013536,
      "grad_norm": 0.4221702218055725,
      "learning_rate": 2.5523139221595556e-06,
      "loss": 0.5574,
      "step": 2740
    },
    {
      "epoch": 0.24169449815433292,
      "grad_norm": 0.3038138449192047,
      "learning_rate": 1.8026057494499821e-06,
      "loss": 0.5472,
      "step": 2750
    },
    {
      "epoch": 0.2425733872385305,
      "grad_norm": 0.3986909091472626,
      "learning_rate": 1.178904772643223e-06,
      "loss": 0.5438,
      "step": 2760
    },
    {
      "epoch": 0.24345227632272806,
      "grad_norm": 0.33036527037620544,
      "learning_rate": 6.90740160441754e-07,
      "loss": 0.5724,
      "step": 2770
    },
    {
      "epoch": 0.24433116540692565,
      "grad_norm": 0.3517259359359741,
      "learning_rate": 3.45570299343632e-07,
      "loss": 0.5737,
      "step": 2780
    },
    {
      "epoch": 0.2452100544911232,
      "grad_norm": 0.388594388961792,
      "learning_rate": 1.486688412425152e-07,
      "loss": 0.5561,
      "step": 2790
    },
    {
      "epoch": 0.24582527685006153,
      "step": 2797,
      "total_flos": 1.2485883762455871e+19,
      "train_loss": 0.05056152754610762,
      "train_runtime": 2891.8532,
      "train_samples_per_second": 11.606,
      "train_steps_per_second": 0.967
    }
  ],
  "logging_steps": 10,
  "max_steps": 2797,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2543,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2485883762455871e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}