{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9410745233968805,
  "eval_steps": 500,
  "global_step": 1120,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0017331022530329288,
      "grad_norm": 647.2084768449807,
      "learning_rate": 0.0,
      "loss": 8.942,
      "step": 1
    },
    {
      "epoch": 0.0034662045060658577,
      "grad_norm": 573.8878102469158,
      "learning_rate": 6.11626355295565e-07,
      "loss": 5.0021,
      "step": 2
    },
    {
      "epoch": 0.005199306759098787,
      "grad_norm": 606.1587044660829,
      "learning_rate": 9.694048375962253e-07,
      "loss": 5.4595,
      "step": 3
    },
    {
      "epoch": 0.006932409012131715,
      "grad_norm": 96.73905784376856,
      "learning_rate": 1.22325271059113e-06,
      "loss": 1.3784,
      "step": 4
    },
    {
      "epoch": 0.008665511265164644,
      "grad_norm": 135.261956495535,
      "learning_rate": 1.4201524179343322e-06,
      "loss": 1.5643,
      "step": 5
    },
    {
      "epoch": 0.010398613518197574,
      "grad_norm": 72.38603313463689,
      "learning_rate": 1.5810311928917904e-06,
      "loss": 1.0616,
      "step": 6
    },
    {
      "epoch": 0.012131715771230503,
      "grad_norm": 110.2572852255691,
      "learning_rate": 1.7170522589991576e-06,
      "loss": 1.5916,
      "step": 7
    },
    {
      "epoch": 0.01386481802426343,
      "grad_norm": 74.340797676772,
      "learning_rate": 1.8348790658866952e-06,
      "loss": 0.7618,
      "step": 8
    },
    {
      "epoch": 0.01559792027729636,
      "grad_norm": 78.43209954018349,
      "learning_rate": 1.9388096751924506e-06,
      "loss": 0.7511,
      "step": 9
    },
    {
      "epoch": 0.01733102253032929,
      "grad_norm": 60.59558927391601,
      "learning_rate": 2.0317787732298976e-06,
      "loss": 0.5852,
      "step": 10
    },
    {
      "epoch": 0.019064124783362217,
      "grad_norm": 43.183686856808265,
      "learning_rate": 2.1158795523013674e-06,
      "loss": 0.631,
      "step": 11
    },
    {
      "epoch": 0.02079722703639515,
      "grad_norm": 31.424876900054738,
      "learning_rate": 2.1926575481873556e-06,
      "loss": 0.5615,
      "step": 12
    },
    {
      "epoch": 0.022530329289428077,
      "grad_norm": 39.9579781101151,
      "learning_rate": 2.2632864577975843e-06,
      "loss": 0.5351,
      "step": 13
    },
    {
      "epoch": 0.024263431542461005,
      "grad_norm": 16.20091369312816,
      "learning_rate": 2.3286786142947225e-06,
      "loss": 0.5437,
      "step": 14
    },
    {
      "epoch": 0.025996533795493933,
      "grad_norm": 21.475372403594005,
      "learning_rate": 2.3895572555305577e-06,
      "loss": 0.3954,
      "step": 15
    },
    {
      "epoch": 0.02772963604852686,
      "grad_norm": 23.317022380974027,
      "learning_rate": 2.44650542118226e-06,
      "loss": 0.5462,
      "step": 16
    },
    {
      "epoch": 0.029462738301559793,
      "grad_norm": 17.029721476707092,
      "learning_rate": 2.5e-06,
      "loss": 0.5015,
      "step": 17
    },
    {
      "epoch": 0.03119584055459272,
      "grad_norm": 27.06097985770979,
      "learning_rate": 2.5504360304880153e-06,
      "loss": 0.5362,
      "step": 18
    },
    {
      "epoch": 0.03292894280762565,
      "grad_norm": 40.86745558242218,
      "learning_rate": 2.598144422607253e-06,
      "loss": 0.5884,
      "step": 19
    },
    {
      "epoch": 0.03466204506065858,
      "grad_norm": 19.042369782246855,
      "learning_rate": 2.6434051285254623e-06,
      "loss": 0.5384,
      "step": 20
    },
    {
      "epoch": 0.036395147313691506,
      "grad_norm": 26.20174950735339,
      "learning_rate": 2.686457096595383e-06,
      "loss": 0.5225,
      "step": 21
    },
    {
      "epoch": 0.038128249566724434,
      "grad_norm": 17.108069299492858,
      "learning_rate": 2.727505907596933e-06,
      "loss": 0.5038,
      "step": 22
    },
    {
      "epoch": 0.03986135181975736,
      "grad_norm": 36.42076561480758,
      "learning_rate": 2.766729712136828e-06,
      "loss": 0.4857,
      "step": 23
    },
    {
      "epoch": 0.0415944540727903,
      "grad_norm": 25.842735630410363,
      "learning_rate": 2.8042839034829207e-06,
      "loss": 0.6238,
      "step": 24
    },
    {
      "epoch": 0.043327556325823226,
      "grad_norm": 14.48908638484767,
      "learning_rate": 2.8403048358686645e-06,
      "loss": 0.489,
      "step": 25
    },
    {
      "epoch": 0.045060658578856154,
      "grad_norm": 22.368583910086443,
      "learning_rate": 2.87491281309315e-06,
      "loss": 0.5438,
      "step": 26
    },
    {
      "epoch": 0.04679376083188908,
      "grad_norm": 23.952114653756958,
      "learning_rate": 2.908214512788676e-06,
      "loss": 0.6398,
      "step": 27
    },
    {
      "epoch": 0.04852686308492201,
      "grad_norm": 30.800322256012496,
      "learning_rate": 2.940304969590288e-06,
      "loss": 0.408,
      "step": 28
    },
    {
      "epoch": 0.05025996533795494,
      "grad_norm": 17.16202657759275,
      "learning_rate": 2.9712692101449995e-06,
      "loss": 0.4687,
      "step": 29
    },
    {
      "epoch": 0.05199306759098787,
      "grad_norm": 29.64958516388762,
      "learning_rate": 3.001183610826123e-06,
      "loss": 0.4357,
      "step": 30
    },
    {
      "epoch": 0.053726169844020795,
      "grad_norm": 23.333477492114202,
      "learning_rate": 3.030117032740661e-06,
      "loss": 0.5406,
      "step": 31
    },
    {
      "epoch": 0.05545927209705372,
      "grad_norm": 15.787395345865024,
      "learning_rate": 3.0581317764778257e-06,
      "loss": 0.5256,
      "step": 32
    },
    {
      "epoch": 0.05719237435008666,
      "grad_norm": 19.45170677561162,
      "learning_rate": 3.0852843898975927e-06,
      "loss": 0.5314,
      "step": 33
    },
    {
      "epoch": 0.058925476603119586,
      "grad_norm": 24.5706801443083,
      "learning_rate": 3.1116263552955654e-06,
      "loss": 0.5072,
      "step": 34
    },
    {
      "epoch": 0.060658578856152515,
      "grad_norm": 29.684724376825955,
      "learning_rate": 3.1372046769334903e-06,
      "loss": 0.4311,
      "step": 35
    },
    {
      "epoch": 0.06239168110918544,
      "grad_norm": 31.311374301434476,
      "learning_rate": 3.162062385783581e-06,
      "loss": 0.5551,
      "step": 36
    },
    {
      "epoch": 0.06412478336221837,
      "grad_norm": 19.19929703379259,
      "learning_rate": 3.1862389751018495e-06,
      "loss": 0.6093,
      "step": 37
    },
    {
      "epoch": 0.0658578856152513,
      "grad_norm": 17.85243347847894,
      "learning_rate": 3.209770777902818e-06,
      "loss": 0.5156,
      "step": 38
    },
    {
      "epoch": 0.06759098786828423,
      "grad_norm": 17.190240371026576,
      "learning_rate": 3.2326912953938096e-06,
      "loss": 0.4895,
      "step": 39
    },
    {
      "epoch": 0.06932409012131716,
      "grad_norm": 33.78868587217843,
      "learning_rate": 3.255031483821028e-06,
      "loss": 0.534,
      "step": 40
    },
    {
      "epoch": 0.07105719237435008,
      "grad_norm": 13.142424724407546,
      "learning_rate": 3.2768200058910072e-06,
      "loss": 0.5617,
      "step": 41
    },
    {
      "epoch": 0.07279029462738301,
      "grad_norm": 17.02905662481777,
      "learning_rate": 3.2980834518909482e-06,
      "loss": 0.5011,
      "step": 42
    },
    {
      "epoch": 0.07452339688041594,
      "grad_norm": 17.419760591983785,
      "learning_rate": 3.3188465347872277e-06,
      "loss": 0.544,
      "step": 43
    },
    {
      "epoch": 0.07625649913344887,
      "grad_norm": 21.31594418078541,
      "learning_rate": 3.3391322628924977e-06,
      "loss": 0.5452,
      "step": 44
    },
    {
      "epoch": 0.0779896013864818,
      "grad_norm": 17.666906007309084,
      "learning_rate": 3.358962093126783e-06,
      "loss": 0.4524,
      "step": 45
    },
    {
      "epoch": 0.07972270363951472,
      "grad_norm": 22.927772226228445,
      "learning_rate": 3.3783560674323935e-06,
      "loss": 0.4701,
      "step": 46
    },
    {
      "epoch": 0.08145580589254767,
      "grad_norm": 30.96923877010864,
      "learning_rate": 3.3973329345169716e-06,
      "loss": 0.56,
      "step": 47
    },
    {
      "epoch": 0.0831889081455806,
      "grad_norm": 33.99353863523644,
      "learning_rate": 3.415910258778486e-06,
      "loss": 0.664,
      "step": 48
    },
    {
      "epoch": 0.08492201039861352,
      "grad_norm": 30.554682506072208,
      "learning_rate": 3.434104517998315e-06,
      "loss": 0.4618,
      "step": 49
    },
    {
      "epoch": 0.08665511265164645,
      "grad_norm": 25.46750859704626,
      "learning_rate": 3.45193119116423e-06,
      "loss": 0.6294,
      "step": 50
    },
    {
      "epoch": 0.08838821490467938,
      "grad_norm": 19.379167268575745,
      "learning_rate": 3.469404837596225e-06,
      "loss": 0.508,
      "step": 51
    },
    {
      "epoch": 0.09012131715771231,
      "grad_norm": 18.243987464731532,
      "learning_rate": 3.486539168388715e-06,
      "loss": 0.4459,
      "step": 52
    },
    {
      "epoch": 0.09185441941074524,
      "grad_norm": 23.016041344672796,
      "learning_rate": 3.503347111047406e-06,
      "loss": 0.4265,
      "step": 53
    },
    {
      "epoch": 0.09358752166377816,
      "grad_norm": 14.495287951701213,
      "learning_rate": 3.519840868084241e-06,
      "loss": 0.6267,
      "step": 54
    },
    {
      "epoch": 0.09532062391681109,
      "grad_norm": 21.848642044293907,
      "learning_rate": 3.5360319702357003e-06,
      "loss": 0.4872,
      "step": 55
    },
    {
      "epoch": 0.09705372616984402,
      "grad_norm": 14.835598090509361,
      "learning_rate": 3.5519313248858533e-06,
      "loss": 0.4769,
      "step": 56
    },
    {
      "epoch": 0.09878682842287695,
      "grad_norm": 28.90002677942639,
      "learning_rate": 3.567549260203478e-06,
      "loss": 0.4979,
      "step": 57
    },
    {
      "epoch": 0.10051993067590988,
      "grad_norm": 10.59256019775053,
      "learning_rate": 3.5828955654405646e-06,
      "loss": 0.47,
      "step": 58
    },
    {
      "epoch": 0.1022530329289428,
      "grad_norm": 31.682652450648664,
      "learning_rate": 3.5979795277859725e-06,
      "loss": 0.5469,
      "step": 59
    },
    {
      "epoch": 0.10398613518197573,
      "grad_norm": 10.631392766048863,
      "learning_rate": 3.6128099661216876e-06,
      "loss": 0.4362,
      "step": 60
    },
    {
      "epoch": 0.10571923743500866,
      "grad_norm": 19.235261898216223,
      "learning_rate": 3.627395261988912e-06,
      "loss": 0.4137,
      "step": 61
    },
    {
      "epoch": 0.10745233968804159,
      "grad_norm": 18.197231140791892,
      "learning_rate": 3.6417433880362255e-06,
      "loss": 0.4565,
      "step": 62
    },
    {
      "epoch": 0.10918544194107452,
      "grad_norm": 16.02045754073376,
      "learning_rate": 3.655861934191608e-06,
      "loss": 0.6155,
      "step": 63
    },
    {
      "epoch": 0.11091854419410745,
      "grad_norm": 16.15375173352492,
      "learning_rate": 3.6697581317733905e-06,
      "loss": 0.4054,
      "step": 64
    },
    {
      "epoch": 0.11265164644714037,
      "grad_norm": 11.412763684873083,
      "learning_rate": 3.6834388757319168e-06,
      "loss": 0.5128,
      "step": 65
    },
    {
      "epoch": 0.11438474870017332,
      "grad_norm": 31.672520574944972,
      "learning_rate": 3.6969107451931574e-06,
      "loss": 0.5469,
      "step": 66
    },
    {
      "epoch": 0.11611785095320624,
      "grad_norm": 23.99710447010678,
      "learning_rate": 3.7101800224575124e-06,
      "loss": 0.5079,
      "step": 67
    },
    {
      "epoch": 0.11785095320623917,
      "grad_norm": 19.960427617020056,
      "learning_rate": 3.7232527105911305e-06,
      "loss": 0.5191,
      "step": 68
    },
    {
      "epoch": 0.1195840554592721,
      "grad_norm": 12.604473082966171,
      "learning_rate": 3.7361345497330536e-06,
      "loss": 0.4771,
      "step": 69
    },
    {
      "epoch": 0.12131715771230503,
      "grad_norm": 13.775862352133236,
      "learning_rate": 3.7488310322290554e-06,
      "loss": 0.4508,
      "step": 70
    },
    {
      "epoch": 0.12305025996533796,
      "grad_norm": 13.777055789803262,
      "learning_rate": 3.7613474166920487e-06,
      "loss": 0.3493,
      "step": 71
    },
    {
      "epoch": 0.12478336221837089,
      "grad_norm": 18.611723193578154,
      "learning_rate": 3.7736887410791456e-06,
      "loss": 0.4584,
      "step": 72
    },
    {
      "epoch": 0.1265164644714038,
      "grad_norm": 13.275033586535018,
      "learning_rate": 3.785859834866764e-06,
      "loss": 0.431,
      "step": 73
    },
    {
      "epoch": 0.12824956672443674,
      "grad_norm": 19.578698614900965,
      "learning_rate": 3.797865330397415e-06,
      "loss": 0.4962,
      "step": 74
    },
    {
      "epoch": 0.12998266897746968,
      "grad_norm": 18.971957744630956,
      "learning_rate": 3.8097096734648898e-06,
      "loss": 0.5146,
      "step": 75
    },
    {
      "epoch": 0.1317157712305026,
      "grad_norm": 27.054719083868367,
      "learning_rate": 3.821397133198383e-06,
      "loss": 0.5843,
      "step": 76
    },
    {
      "epoch": 0.13344887348353554,
      "grad_norm": 12.07436092099249,
      "learning_rate": 3.832931811300525e-06,
      "loss": 0.445,
      "step": 77
    },
    {
      "epoch": 0.13518197573656845,
      "grad_norm": 18.498393220974965,
      "learning_rate": 3.844317650689375e-06,
      "loss": 0.505,
      "step": 78
    },
    {
      "epoch": 0.1369150779896014,
      "grad_norm": 9.554232744824942,
      "learning_rate": 3.855558443589912e-06,
      "loss": 0.3914,
      "step": 79
    },
    {
      "epoch": 0.1386481802426343,
      "grad_norm": 20.695634700445815,
      "learning_rate": 3.866657839116593e-06,
      "loss": 0.4728,
      "step": 80
    },
    {
      "epoch": 0.14038128249566725,
      "grad_norm": 23.70650056303782,
      "learning_rate": 3.877619350384901e-06,
      "loss": 0.4131,
      "step": 81
    },
    {
      "epoch": 0.14211438474870017,
      "grad_norm": 17.362607639492655,
      "learning_rate": 3.888446361186572e-06,
      "loss": 0.5666,
      "step": 82
    },
    {
      "epoch": 0.1438474870017331,
      "grad_norm": 30.803312405905007,
      "learning_rate": 3.899142132260232e-06,
      "loss": 0.3942,
      "step": 83
    },
    {
      "epoch": 0.14558058925476602,
      "grad_norm": 14.433917240962181,
      "learning_rate": 3.909709807186513e-06,
      "loss": 0.3539,
      "step": 84
    },
    {
      "epoch": 0.14731369150779897,
      "grad_norm": 27.983714135831534,
      "learning_rate": 3.920152417934333e-06,
      "loss": 0.4866,
      "step": 85
    },
    {
      "epoch": 0.14904679376083188,
      "grad_norm": 27.656373587294485,
      "learning_rate": 3.930472890082793e-06,
      "loss": 0.6394,
      "step": 86
    },
    {
      "epoch": 0.15077989601386482,
      "grad_norm": 20.297289226671918,
      "learning_rate": 3.940674047741225e-06,
      "loss": 0.2724,
      "step": 87
    },
    {
      "epoch": 0.15251299826689774,
      "grad_norm": 17.38939306620382,
      "learning_rate": 3.950758618188063e-06,
      "loss": 0.3958,
      "step": 88
    },
    {
      "epoch": 0.15424610051993068,
      "grad_norm": 23.3167566908034,
      "learning_rate": 3.960729236247623e-06,
      "loss": 0.4648,
      "step": 89
    },
    {
      "epoch": 0.1559792027729636,
      "grad_norm": 14.519958965989657,
      "learning_rate": 3.970588448422349e-06,
      "loss": 0.4619,
      "step": 90
    },
    {
      "epoch": 0.15771230502599654,
      "grad_norm": 16.291955865633028,
      "learning_rate": 3.980338716796741e-06,
      "loss": 0.4586,
      "step": 91
    },
    {
      "epoch": 0.15944540727902945,
      "grad_norm": 11.652366011544274,
      "learning_rate": 3.989982422727958e-06,
      "loss": 0.4995,
      "step": 92
    },
    {
      "epoch": 0.1611785095320624,
      "grad_norm": 6.2747635633447505,
      "learning_rate": 3.9995218703368865e-06,
      "loss": 0.4076,
      "step": 93
    },
    {
      "epoch": 0.16291161178509533,
      "grad_norm": 12.44265839460381,
      "learning_rate": 4.008959289812537e-06,
      "loss": 0.5292,
      "step": 94
    },
    {
      "epoch": 0.16464471403812825,
      "grad_norm": 8.095963458947507,
      "learning_rate": 4.018296840541585e-06,
      "loss": 0.5759,
      "step": 95
    },
    {
      "epoch": 0.1663778162911612,
      "grad_norm": 11.188171707532316,
      "learning_rate": 4.027536614074051e-06,
      "loss": 0.4963,
      "step": 96
    },
    {
      "epoch": 0.1681109185441941,
      "grad_norm": 11.513044836880278,
      "learning_rate": 4.036680636935308e-06,
      "loss": 0.5485,
      "step": 97
    },
    {
      "epoch": 0.16984402079722705,
      "grad_norm": 7.033277973602002,
      "learning_rate": 4.045730873293881e-06,
      "loss": 0.4133,
      "step": 98
    },
    {
      "epoch": 0.17157712305025996,
      "grad_norm": 11.474249495686323,
      "learning_rate": 4.054689227493818e-06,
      "loss": 0.5028,
      "step": 99
    },
    {
      "epoch": 0.1733102253032929,
      "grad_norm": 6.876384526337422,
      "learning_rate": 4.063557546459795e-06,
      "loss": 0.4684,
      "step": 100
    },
    {
      "epoch": 0.17504332755632582,
      "grad_norm": 5.545370752238767,
      "learning_rate": 4.072337621982561e-06,
      "loss": 0.3779,
      "step": 101
    },
    {
      "epoch": 0.17677642980935876,
      "grad_norm": 10.45499475290175,
      "learning_rate": 4.08103119289179e-06,
      "loss": 0.4031,
      "step": 102
    },
    {
      "epoch": 0.17850953206239167,
      "grad_norm": 6.615622739651834,
      "learning_rate": 4.089639947122947e-06,
      "loss": 0.3705,
      "step": 103
    },
    {
      "epoch": 0.18024263431542462,
      "grad_norm": 17.496175293541835,
      "learning_rate": 4.09816552368428e-06,
      "loss": 0.4729,
      "step": 104
    },
    {
      "epoch": 0.18197573656845753,
      "grad_norm": 18.752328143220392,
      "learning_rate": 4.106609514529715e-06,
      "loss": 0.5593,
      "step": 105
    },
    {
      "epoch": 0.18370883882149047,
      "grad_norm": 7.70506252968954,
      "learning_rate": 4.114973466342971e-06,
      "loss": 0.2386,
      "step": 106
    },
    {
      "epoch": 0.1854419410745234,
      "grad_norm": 35.3334423519348,
      "learning_rate": 4.12325888223791e-06,
      "loss": 0.5985,
      "step": 107
    },
    {
      "epoch": 0.18717504332755633,
      "grad_norm": 13.458069852881543,
      "learning_rate": 4.1314672233798066e-06,
      "loss": 0.5091,
      "step": 108
    },
    {
      "epoch": 0.18890814558058924,
      "grad_norm": 14.002242317464448,
      "learning_rate": 4.139599910531887e-06,
      "loss": 0.5079,
      "step": 109
    },
    {
      "epoch": 0.19064124783362218,
      "grad_norm": 15.454545344388421,
      "learning_rate": 4.1476583255312654e-06,
      "loss": 0.6272,
      "step": 110
    },
    {
      "epoch": 0.1923743500866551,
      "grad_norm": 9.983712092258735,
      "learning_rate": 4.155643812698074e-06,
      "loss": 0.4514,
      "step": 111
    },
    {
      "epoch": 0.19410745233968804,
      "grad_norm": 27.964194155833567,
      "learning_rate": 4.1635576801814176e-06,
      "loss": 0.5969,
      "step": 112
    },
    {
      "epoch": 0.19584055459272098,
      "grad_norm": 12.856476430424266,
      "learning_rate": 4.1714012012454905e-06,
      "loss": 0.4496,
      "step": 113
    },
    {
      "epoch": 0.1975736568457539,
      "grad_norm": 11.994856049793738,
      "learning_rate": 4.179175615499044e-06,
      "loss": 0.4286,
      "step": 114
    },
    {
      "epoch": 0.19930675909878684,
      "grad_norm": 8.955000809031013,
      "learning_rate": 4.186882130071161e-06,
      "loss": 0.4467,
      "step": 115
    },
    {
      "epoch": 0.20103986135181975,
      "grad_norm": 7.105558674784872,
      "learning_rate": 4.194521920736129e-06,
      "loss": 0.5011,
      "step": 116
    },
    {
      "epoch": 0.2027729636048527,
      "grad_norm": 13.637778353276092,
      "learning_rate": 4.202096132990035e-06,
      "loss": 0.5012,
      "step": 117
    },
    {
      "epoch": 0.2045060658578856,
      "grad_norm": 10.984593092162797,
      "learning_rate": 4.209605883081538e-06,
      "loss": 0.4904,
      "step": 118
    },
    {
      "epoch": 0.20623916811091855,
      "grad_norm": 7.149941408144026,
      "learning_rate": 4.2170522589991584e-06,
      "loss": 0.3922,
      "step": 119
    },
    {
      "epoch": 0.20797227036395147,
      "grad_norm": 7.360129153735549,
      "learning_rate": 4.224436321417253e-06,
      "loss": 0.3727,
      "step": 120
    },
    {
      "epoch": 0.2097053726169844,
      "grad_norm": 6.961896733939193,
      "learning_rate": 4.231759104602735e-06,
      "loss": 0.3628,
      "step": 121
    },
    {
      "epoch": 0.21143847487001732,
      "grad_norm": 10.700167816842582,
      "learning_rate": 4.239021617284477e-06,
      "loss": 0.4458,
      "step": 122
    },
    {
      "epoch": 0.21317157712305027,
      "grad_norm": 23.11925340337978,
      "learning_rate": 4.2462248434872325e-06,
      "loss": 0.5965,
      "step": 123
    },
    {
      "epoch": 0.21490467937608318,
      "grad_norm": 20.38587259033891,
      "learning_rate": 4.253369743331791e-06,
      "loss": 0.4362,
      "step": 124
    },
    {
      "epoch": 0.21663778162911612,
      "grad_norm": 18.568964901672125,
      "learning_rate": 4.260457253802998e-06,
      "loss": 0.4779,
      "step": 125
    },
    {
      "epoch": 0.21837088388214904,
      "grad_norm": 18.589908183724223,
      "learning_rate": 4.267488289487173e-06,
      "loss": 0.515,
      "step": 126
    },
    {
      "epoch": 0.22010398613518198,
      "grad_norm": 16.449700646734808,
      "learning_rate": 4.274463743280388e-06,
      "loss": 0.4853,
      "step": 127
    },
    {
      "epoch": 0.2218370883882149,
      "grad_norm": 12.18118263253402,
      "learning_rate": 4.281384487068956e-06,
      "loss": 0.3953,
      "step": 128
    },
    {
      "epoch": 0.22357019064124783,
      "grad_norm": 8.802206317712072,
      "learning_rate": 4.288251372383453e-06,
      "loss": 0.3602,
      "step": 129
    },
    {
      "epoch": 0.22530329289428075,
      "grad_norm": 9.476806033685488,
      "learning_rate": 4.295065231027482e-06,
      "loss": 0.4143,
      "step": 130
    },
    {
      "epoch": 0.2270363951473137,
      "grad_norm": 14.948175039506285,
      "learning_rate": 4.3018268756823446e-06,
      "loss": 0.371,
      "step": 131
    },
    {
      "epoch": 0.22876949740034663,
      "grad_norm": 21.025966757230734,
      "learning_rate": 4.308537100488723e-06,
      "loss": 0.6845,
      "step": 132
    },
    {
      "epoch": 0.23050259965337955,
      "grad_norm": 12.536131861574416,
      "learning_rate": 4.31519668160641e-06,
      "loss": 0.4567,
      "step": 133
    },
    {
      "epoch": 0.2322357019064125,
      "grad_norm": 7.07413732438644,
      "learning_rate": 4.321806377753077e-06,
      "loss": 0.4183,
      "step": 134
    },
    {
      "epoch": 0.2339688041594454,
      "grad_norm": 14.003418641266785,
      "learning_rate": 4.328366930723009e-06,
      "loss": 0.4637,
      "step": 135
    },
    {
      "epoch": 0.23570190641247835,
      "grad_norm": 19.47673955745834,
      "learning_rate": 4.334879065886696e-06,
      "loss": 0.4894,
      "step": 136
    },
    {
      "epoch": 0.23743500866551126,
      "grad_norm": 13.83455061248518,
      "learning_rate": 4.341343492672135e-06,
      "loss": 0.4255,
      "step": 137
    },
    {
      "epoch": 0.2391681109185442,
      "grad_norm": 10.549835015765094,
      "learning_rate": 4.347760905028619e-06,
      "loss": 0.3705,
      "step": 138
    },
    {
      "epoch": 0.24090121317157712,
      "grad_norm": 19.880365763010843,
      "learning_rate": 4.354131981873779e-06,
      "loss": 0.463,
      "step": 139
    },
    {
      "epoch": 0.24263431542461006,
      "grad_norm": 8.317820287694301,
      "learning_rate": 4.36045738752462e-06,
      "loss": 0.4416,
      "step": 140
    },
    {
      "epoch": 0.24436741767764297,
      "grad_norm": 13.40399791234614,
      "learning_rate": 4.366737772113197e-06,
      "loss": 0.472,
      "step": 141
    },
    {
      "epoch": 0.24610051993067592,
      "grad_norm": 8.528376168605732,
      "learning_rate": 4.372973771987614e-06,
      "loss": 0.5117,
      "step": 142
    },
    {
      "epoch": 0.24783362218370883,
      "grad_norm": 9.354057418989575,
      "learning_rate": 4.379166010098952e-06,
      "loss": 0.5081,
      "step": 143
    },
    {
      "epoch": 0.24956672443674177,
      "grad_norm": 11.444925201424907,
      "learning_rate": 4.385315096374711e-06,
      "loss": 0.4803,
      "step": 144
    },
    {
      "epoch": 0.2512998266897747,
      "grad_norm": 19.89077689209483,
      "learning_rate": 4.391421628079332e-06,
      "loss": 0.4154,
      "step": 145
    },
    {
      "epoch": 0.2530329289428076,
      "grad_norm": 24.657033465451725,
      "learning_rate": 4.397486190162329e-06,
      "loss": 0.5379,
      "step": 146
    },
    {
      "epoch": 0.25476603119584057,
      "grad_norm": 12.398974690040214,
      "learning_rate": 4.4035093555945405e-06,
      "loss": 0.4077,
      "step": 147
    },
    {
      "epoch": 0.2564991334488735,
      "grad_norm": 9.179246997409598,
      "learning_rate": 4.409491685692979e-06,
      "loss": 0.428,
      "step": 148
    },
    {
      "epoch": 0.2582322357019064,
      "grad_norm": 17.731797031789036,
      "learning_rate": 4.415433730434749e-06,
      "loss": 0.5006,
      "step": 149
    },
    {
      "epoch": 0.25996533795493937,
      "grad_norm": 11.24828986839239,
      "learning_rate": 4.421336028760455e-06,
      "loss": 0.4115,
      "step": 150
    },
    {
      "epoch": 0.2616984402079723,
      "grad_norm": 16.050868234888416,
      "learning_rate": 4.427199108867543e-06,
      "loss": 0.4695,
      "step": 151
    },
    {
      "epoch": 0.2634315424610052,
      "grad_norm": 15.9121500101129,
      "learning_rate": 4.433023488493948e-06,
      "loss": 0.4928,
      "step": 152
    },
    {
      "epoch": 0.2651646447140381,
      "grad_norm": 8.656817586310586,
      "learning_rate": 4.43880967519245e-06,
      "loss": 0.3921,
      "step": 153
    },
    {
      "epoch": 0.2668977469670711,
      "grad_norm": 15.08556376171567,
      "learning_rate": 4.444558166596091e-06,
      "loss": 0.3893,
      "step": 154
    },
    {
      "epoch": 0.268630849220104,
      "grad_norm": 9.548275409384516,
      "learning_rate": 4.450269450674993e-06,
      "loss": 0.364,
      "step": 155
    },
    {
      "epoch": 0.2703639514731369,
      "grad_norm": 13.23211787588991,
      "learning_rate": 4.45594400598494e-06,
      "loss": 0.4719,
      "step": 156
    },
    {
      "epoch": 0.2720970537261698,
      "grad_norm": 11.609559007626538,
      "learning_rate": 4.461582301907991e-06,
      "loss": 0.428,
      "step": 157
    },
    {
      "epoch": 0.2738301559792028,
      "grad_norm": 9.871504925646864,
      "learning_rate": 4.4671847988854775e-06,
      "loss": 0.3883,
      "step": 158
    },
    {
      "epoch": 0.2755632582322357,
      "grad_norm": 15.764460147285462,
      "learning_rate": 4.472751948643632e-06,
      "loss": 0.3507,
      "step": 159
    },
    {
      "epoch": 0.2772963604852686,
      "grad_norm": 12.767257679405697,
      "learning_rate": 4.478284194412158e-06,
      "loss": 0.3033,
      "step": 160
    },
    {
      "epoch": 0.27902946273830154,
      "grad_norm": 12.690860397233527,
      "learning_rate": 4.483781971135986e-06,
      "loss": 0.4317,
      "step": 161
    },
    {
      "epoch": 0.2807625649913345,
      "grad_norm": 22.42458607303321,
      "learning_rate": 4.489245705680466e-06,
      "loss": 0.4554,
      "step": 162
    },
    {
      "epoch": 0.2824956672443674,
      "grad_norm": 19.714364266880775,
      "learning_rate": 4.49467581703026e-06,
      "loss": 0.4623,
      "step": 163
    },
    {
      "epoch": 0.28422876949740034,
      "grad_norm": 20.227522820240726,
      "learning_rate": 4.500072716482138e-06,
      "loss": 0.4792,
      "step": 164
    },
    {
      "epoch": 0.28596187175043325,
      "grad_norm": 9.147591829508839,
      "learning_rate": 4.505436807831925e-06,
      "loss": 0.3672,
      "step": 165
    },
    {
      "epoch": 0.2876949740034662,
      "grad_norm": 10.088070053764891,
      "learning_rate": 4.510768487555797e-06,
      "loss": 0.3504,
      "step": 166
    },
    {
      "epoch": 0.28942807625649913,
      "grad_norm": 14.5308121153782,
      "learning_rate": 4.5160681449861246e-06,
      "loss": 0.4113,
      "step": 167
    },
    {
      "epoch": 0.29116117850953205,
      "grad_norm": 11.169616727067668,
      "learning_rate": 4.5213361624820785e-06,
      "loss": 0.4504,
      "step": 168
    },
    {
      "epoch": 0.292894280762565,
      "grad_norm": 8.389957110144163,
      "learning_rate": 4.526572915595169e-06,
      "loss": 0.4449,
      "step": 169
    },
    {
      "epoch": 0.29462738301559793,
      "grad_norm": 15.010197034504525,
      "learning_rate": 4.531778773229898e-06,
      "loss": 0.5587,
      "step": 170
    },
    {
      "epoch": 0.29636048526863085,
      "grad_norm": 10.21208286063813,
      "learning_rate": 4.536954097799704e-06,
      "loss": 0.4414,
      "step": 171
    },
    {
      "epoch": 0.29809358752166376,
      "grad_norm": 12.638185772650418,
      "learning_rate": 4.542099245378358e-06,
      "loss": 0.5566,
      "step": 172
    },
    {
      "epoch": 0.29982668977469673,
      "grad_norm": 8.705210245870267,
      "learning_rate": 4.5472145658469775e-06,
      "loss": 0.4252,
      "step": 173
    },
    {
      "epoch": 0.30155979202772965,
      "grad_norm": 11.171343341261151,
      "learning_rate": 4.55230040303679e-06,
      "loss": 0.4217,
      "step": 174
    },
    {
      "epoch": 0.30329289428076256,
      "grad_norm": 16.558480865532886,
      "learning_rate": 4.557357094867823e-06,
      "loss": 0.3367,
      "step": 175
    },
    {
      "epoch": 0.3050259965337955,
      "grad_norm": 6.442331519449627,
      "learning_rate": 4.562384973483628e-06,
      "loss": 0.5464,
      "step": 176
    },
    {
      "epoch": 0.30675909878682844,
      "grad_norm": 10.451564792125813,
      "learning_rate": 4.567384365382197e-06,
      "loss": 0.4271,
      "step": 177
    },
    {
      "epoch": 0.30849220103986136,
      "grad_norm": 10.277068705341831,
      "learning_rate": 4.572355591543188e-06,
      "loss": 0.3607,
      "step": 178
    },
    {
      "epoch": 0.31022530329289427,
      "grad_norm": 13.429148363051022,
      "learning_rate": 4.577298967551584e-06,
      "loss": 0.4127,
      "step": 179
    },
    {
      "epoch": 0.3119584055459272,
      "grad_norm": 7.501831085475716,
      "learning_rate": 4.582214803717913e-06,
      "loss": 0.3702,
      "step": 180
    },
    {
      "epoch": 0.31369150779896016,
      "grad_norm": 4.787893429519045,
      "learning_rate": 4.587103405195136e-06,
      "loss": 0.4849,
      "step": 181
    },
    {
      "epoch": 0.31542461005199307,
      "grad_norm": 6.992617392510967,
      "learning_rate": 4.591965072092307e-06,
      "loss": 0.4492,
      "step": 182
    },
    {
      "epoch": 0.317157712305026,
      "grad_norm": 14.9714297004458,
      "learning_rate": 4.596800099585137e-06,
      "loss": 0.454,
      "step": 183
    },
    {
      "epoch": 0.3188908145580589,
      "grad_norm": 15.480837882774123,
      "learning_rate": 4.601608778023523e-06,
      "loss": 0.5298,
      "step": 184
    },
    {
      "epoch": 0.32062391681109187,
      "grad_norm": 9.95944803508579,
      "learning_rate": 4.606391393036181e-06,
      "loss": 0.4223,
      "step": 185
    },
    {
      "epoch": 0.3223570190641248,
      "grad_norm": 12.949212858218871,
      "learning_rate": 4.611148225632451e-06,
      "loss": 0.3895,
      "step": 186
    },
    {
      "epoch": 0.3240901213171577,
      "grad_norm": 5.746041967390738,
      "learning_rate": 4.615879552301368e-06,
      "loss": 0.3128,
      "step": 187
    },
    {
      "epoch": 0.32582322357019067,
      "grad_norm": 13.874993951469683,
      "learning_rate": 4.620585645108102e-06,
      "loss": 0.5579,
      "step": 188
    },
    {
      "epoch": 0.3275563258232236,
      "grad_norm": 6.919014987850482,
      "learning_rate": 4.625266771787833e-06,
      "loss": 0.4351,
      "step": 189
    },
    {
      "epoch": 0.3292894280762565,
      "grad_norm": 9.669997814221613,
      "learning_rate": 4.6299231958371505e-06,
      "loss": 0.492,
      "step": 190
    },
    {
      "epoch": 0.3310225303292894,
      "grad_norm": 12.28345141688351,
      "learning_rate": 4.634555176603051e-06,
      "loss": 0.4062,
      "step": 191
    },
    {
      "epoch": 0.3327556325823224,
      "grad_norm": 12.274093289887738,
      "learning_rate": 4.639162969369616e-06,
      "loss": 0.4863,
      "step": 192
    },
    {
      "epoch": 0.3344887348353553,
      "grad_norm": 12.697327364149624,
      "learning_rate": 4.643746825442441e-06,
      "loss": 0.4765,
      "step": 193
    },
    {
      "epoch": 0.3362218370883882,
      "grad_norm": 6.233879968402441,
      "learning_rate": 4.648306992230872e-06,
      "loss": 0.4208,
      "step": 194
    },
    {
      "epoch": 0.3379549393414211,
      "grad_norm": 16.365763847690072,
      "learning_rate": 4.6528437133281425e-06,
      "loss": 0.4297,
      "step": 195
    },
    {
      "epoch": 0.3396880415944541,
      "grad_norm": 5.830781469299343,
      "learning_rate": 4.657357228589445e-06,
      "loss": 0.4274,
      "step": 196
    },
    {
      "epoch": 0.341421143847487,
      "grad_norm": 12.279480008227585,
      "learning_rate": 4.6618477742080345e-06,
      "loss": 0.3753,
      "step": 197
    },
    {
      "epoch": 0.3431542461005199,
      "grad_norm": 14.330985712408543,
      "learning_rate": 4.666315582789383e-06,
      "loss": 0.3959,
      "step": 198
    },
    {
      "epoch": 0.34488734835355284,
      "grad_norm": 8.411018919121066,
      "learning_rate": 4.670760883423489e-06,
      "loss": 0.3989,
      "step": 199
    },
    {
      "epoch": 0.3466204506065858,
      "grad_norm": 13.321687635213314,
      "learning_rate": 4.67518390175536e-06,
      "loss": 0.4854,
      "step": 200
    },
    {
      "epoch": 0.3483535528596187,
      "grad_norm": 7.844456857389884,
      "learning_rate": 4.679584860053737e-06,
      "loss": 0.4072,
      "step": 201
    },
    {
      "epoch": 0.35008665511265163,
      "grad_norm": 4.942686809121776,
      "learning_rate": 4.683963977278126e-06,
      "loss": 0.3907,
      "step": 202
    },
    {
      "epoch": 0.35181975736568455,
      "grad_norm": 9.700514585079084,
      "learning_rate": 4.688321469144157e-06,
      "loss": 0.4527,
      "step": 203
    },
    {
      "epoch": 0.3535528596187175,
      "grad_norm": 15.641539098225856,
      "learning_rate": 4.692657548187356e-06,
      "loss": 0.4047,
      "step": 204
    },
    {
      "epoch": 0.35528596187175043,
      "grad_norm": 8.40988727144764,
      "learning_rate": 4.69697242382534e-06,
      "loss": 0.514,
      "step": 205
    },
    {
      "epoch": 0.35701906412478335,
      "grad_norm": 5.893269479040296,
      "learning_rate": 4.7012663024185115e-06,
      "loss": 0.3671,
      "step": 206
    },
    {
      "epoch": 0.3587521663778163,
      "grad_norm": 6.561880609255776,
      "learning_rate": 4.705539387329278e-06,
      "loss": 0.2395,
      "step": 207
    },
    {
      "epoch": 0.36048526863084923,
      "grad_norm": 9.631120197428968,
      "learning_rate": 4.709791878979845e-06,
      "loss": 0.4619,
      "step": 208
    },
    {
      "epoch": 0.36221837088388215,
      "grad_norm": 9.499252289911182,
      "learning_rate": 4.71402397490862e-06,
      "loss": 0.3724,
      "step": 209
    },
    {
      "epoch": 0.36395147313691506,
      "grad_norm": 11.264829906495066,
      "learning_rate": 4.718235869825281e-06,
      "loss": 0.4255,
      "step": 210
    },
    {
      "epoch": 0.36568457538994803,
      "grad_norm": 18.494331934966,
      "learning_rate": 4.72242775566452e-06,
      "loss": 0.4323,
      "step": 211
    },
    {
      "epoch": 0.36741767764298094,
      "grad_norm": 18.6571872801618,
      "learning_rate": 4.7265998216385365e-06,
      "loss": 0.649,
      "step": 212
    },
    {
      "epoch": 0.36915077989601386,
      "grad_norm": 12.503936758007143,
      "learning_rate": 4.7307522542882744e-06,
      "loss": 0.4432,
      "step": 213
    },
    {
      "epoch": 0.3708838821490468,
      "grad_norm": 8.47141575822663,
      "learning_rate": 4.734885237533476e-06,
      "loss": 0.392,
      "step": 214
    },
    {
      "epoch": 0.37261698440207974,
      "grad_norm": 10.61706929144552,
      "learning_rate": 4.7389989527215606e-06,
      "loss": 0.4142,
      "step": 215
    },
    {
      "epoch": 0.37435008665511266,
      "grad_norm": 9.014027154766897,
      "learning_rate": 4.743093578675371e-06,
      "loss": 0.4947,
      "step": 216
    },
    {
      "epoch": 0.37608318890814557,
      "grad_norm": 8.320984146159908,
      "learning_rate": 4.747169291739819e-06,
      "loss": 0.3772,
      "step": 217
    },
    {
      "epoch": 0.3778162911611785,
      "grad_norm": 8.53816362206238,
      "learning_rate": 4.7512262658274515e-06,
      "loss": 0.3292,
      "step": 218
    },
    {
      "epoch": 0.37954939341421146,
      "grad_norm": 14.132431370966502,
      "learning_rate": 4.755264672462989e-06,
      "loss": 0.4386,
      "step": 219
    },
    {
      "epoch": 0.38128249566724437,
      "grad_norm": 9.175806194091255,
      "learning_rate": 4.75928468082683e-06,
      "loss": 0.3848,
      "step": 220
    },
    {
      "epoch": 0.3830155979202773,
      "grad_norm": 15.096486351018385,
      "learning_rate": 4.763286457797585e-06,
      "loss": 0.5215,
      "step": 221
    },
    {
      "epoch": 0.3847487001733102,
      "grad_norm": 8.551266617677264,
      "learning_rate": 4.76727016799364e-06,
      "loss": 0.3721,
      "step": 222
    },
    {
      "epoch": 0.38648180242634317,
      "grad_norm": 13.28566677477614,
      "learning_rate": 4.771235973813795e-06,
      "loss": 0.5972,
      "step": 223
    },
    {
      "epoch": 0.3882149046793761,
      "grad_norm": 12.3087016021933,
      "learning_rate": 4.775184035476983e-06,
      "loss": 0.4653,
      "step": 224
    },
    {
      "epoch": 0.389948006932409,
      "grad_norm": 7.15832649322408,
      "learning_rate": 4.7791145110611155e-06,
      "loss": 0.2707,
      "step": 225
    },
    {
      "epoch": 0.39168110918544197,
      "grad_norm": 9.170227838678887,
      "learning_rate": 4.783027556541055e-06,
      "loss": 0.4526,
      "step": 226
    },
    {
      "epoch": 0.3934142114384749,
      "grad_norm": 8.37008705084777,
      "learning_rate": 4.786923325825761e-06,
      "loss": 0.383,
      "step": 227
    },
    {
      "epoch": 0.3951473136915078,
      "grad_norm": 11.555910944568112,
      "learning_rate": 4.7908019707946085e-06,
      "loss": 0.4673,
      "step": 228
    },
    {
      "epoch": 0.3968804159445407,
      "grad_norm": 16.504061112439114,
      "learning_rate": 4.794663641332922e-06,
      "loss": 0.4911,
      "step": 229
    },
    {
      "epoch": 0.3986135181975737,
      "grad_norm": 14.071885431409914,
      "learning_rate": 4.7985084853667255e-06,
      "loss": 0.4561,
      "step": 230
    },
    {
      "epoch": 0.4003466204506066,
      "grad_norm": 6.285698828740126,
      "learning_rate": 4.80233664889675e-06,
      "loss": 0.443,
      "step": 231
    },
    {
      "epoch": 0.4020797227036395,
      "grad_norm": 13.229149301272232,
      "learning_rate": 4.806148276031695e-06,
      "loss": 0.3523,
      "step": 232
    },
    {
      "epoch": 0.4038128249566724,
      "grad_norm": 13.81591368043371,
      "learning_rate": 4.809943509020779e-06,
      "loss": 0.4183,
      "step": 233
    },
    {
      "epoch": 0.4055459272097054,
      "grad_norm": 16.47086883754685,
      "learning_rate": 4.8137224882856005e-06,
      "loss": 0.5456,
      "step": 234
    },
    {
      "epoch": 0.4072790294627383,
      "grad_norm": 18.328050904315273,
      "learning_rate": 4.8174853524513045e-06,
      "loss": 0.5196,
      "step": 235
    },
    {
      "epoch": 0.4090121317157712,
      "grad_norm": 10.680748664246744,
      "learning_rate": 4.821232238377102e-06,
      "loss": 0.3868,
      "step": 236
    },
    {
      "epoch": 0.41074523396880414,
      "grad_norm": 8.204132192631935,
      "learning_rate": 4.824963281186138e-06,
      "loss": 0.4306,
      "step": 237
    },
    {
      "epoch": 0.4124783362218371,
      "grad_norm": 14.892637423037684,
      "learning_rate": 4.828678614294723e-06,
      "loss": 0.4233,
      "step": 238
    },
    {
      "epoch": 0.41421143847487,
      "grad_norm": 8.709528158132603,
      "learning_rate": 4.832378369440971e-06,
      "loss": 0.2982,
      "step": 239
    },
    {
      "epoch": 0.41594454072790293,
      "grad_norm": 5.958965170200125,
      "learning_rate": 4.836062676712818e-06,
      "loss": 0.4226,
      "step": 240
    },
    {
      "epoch": 0.41767764298093585,
      "grad_norm": 10.24467627740421,
      "learning_rate": 4.839731664575475e-06,
      "loss": 0.3496,
      "step": 241
    },
    {
      "epoch": 0.4194107452339688,
      "grad_norm": 5.539483790120768,
      "learning_rate": 4.8433854598983e-06,
      "loss": 0.3415,
      "step": 242
    },
    {
      "epoch": 0.42114384748700173,
      "grad_norm": 8.128522373504454,
      "learning_rate": 4.847024187981126e-06,
      "loss": 0.5147,
      "step": 243
    },
    {
      "epoch": 0.42287694974003465,
      "grad_norm": 9.820921958638904,
      "learning_rate": 4.850647972580042e-06,
      "loss": 0.3599,
      "step": 244
    },
    {
      "epoch": 0.4246100519930676,
      "grad_norm": 10.176041413420853,
      "learning_rate": 4.854256935932648e-06,
      "loss": 0.3965,
      "step": 245
    },
    {
      "epoch": 0.42634315424610053,
      "grad_norm": 13.707912326520129,
      "learning_rate": 4.857851198782797e-06,
      "loss": 0.5322,
      "step": 246
    },
    {
      "epoch": 0.42807625649913345,
      "grad_norm": 11.441873016682313,
      "learning_rate": 4.861430880404838e-06,
      "loss": 0.378,
      "step": 247
    },
    {
      "epoch": 0.42980935875216636,
      "grad_norm": 6.501877733585891,
      "learning_rate": 4.864996098627357e-06,
      "loss": 0.461,
      "step": 248
    },
    {
      "epoch": 0.43154246100519933,
      "grad_norm": 8.929733636691838,
      "learning_rate": 4.868546969856457e-06,
      "loss": 0.4073,
      "step": 249
    },
    {
      "epoch": 0.43327556325823224,
      "grad_norm": 7.94527240994372,
      "learning_rate": 4.872083609098562e-06,
      "loss": 0.3213,
      "step": 250
    },
    {
      "epoch": 0.43500866551126516,
      "grad_norm": 14.798819231924542,
      "learning_rate": 4.875606129982766e-06,
      "loss": 0.5674,
      "step": 251
    },
    {
      "epoch": 0.43674176776429807,
      "grad_norm": 15.105036904707585,
      "learning_rate": 4.879114644782739e-06,
      "loss": 0.5448,
      "step": 252
    },
    {
      "epoch": 0.43847487001733104,
      "grad_norm": 12.087473894345516,
      "learning_rate": 4.882609264438196e-06,
      "loss": 0.4241,
      "step": 253
    },
    {
      "epoch": 0.44020797227036396,
      "grad_norm": 19.071727883004186,
      "learning_rate": 4.886090098575953e-06,
      "loss": 0.4046,
      "step": 254
    },
    {
      "epoch": 0.44194107452339687,
      "grad_norm": 12.372767391525821,
      "learning_rate": 4.889557255530558e-06,
      "loss": 0.5612,
      "step": 255
    },
    {
      "epoch": 0.4436741767764298,
      "grad_norm": 10.50741133934516,
      "learning_rate": 4.89301084236452e-06,
      "loss": 0.5331,
      "step": 256
    },
    {
      "epoch": 0.44540727902946275,
      "grad_norm": 19.316048009190684,
      "learning_rate": 4.896450964888154e-06,
      "loss": 0.5252,
      "step": 257
    },
    {
      "epoch": 0.44714038128249567,
      "grad_norm": 15.641739080576807,
      "learning_rate": 4.8998777276790185e-06,
      "loss": 0.4312,
      "step": 258
    },
    {
      "epoch": 0.4488734835355286,
      "grad_norm": 15.977443530872721,
      "learning_rate": 4.903291234101007e-06,
      "loss": 0.5486,
      "step": 259
    },
    {
      "epoch": 0.4506065857885615,
      "grad_norm": 17.842888004573418,
      "learning_rate": 4.906691586323047e-06,
      "loss": 0.5822,
      "step": 260
    },
    {
      "epoch": 0.45233968804159447,
      "grad_norm": 8.564677831644532,
      "learning_rate": 4.9100788853374505e-06,
      "loss": 0.4063,
      "step": 261
    },
    {
      "epoch": 0.4540727902946274,
      "grad_norm": 7.905443715835202,
      "learning_rate": 4.913453230977909e-06,
      "loss": 0.3823,
      "step": 262
    },
    {
      "epoch": 0.4558058925476603,
      "grad_norm": 6.231718869490132,
      "learning_rate": 4.916814721937159e-06,
      "loss": 0.4988,
      "step": 263
    },
    {
      "epoch": 0.45753899480069327,
      "grad_norm": 9.990148269118174,
      "learning_rate": 4.920163455784288e-06,
      "loss": 0.3473,
      "step": 264
    },
    {
      "epoch": 0.4592720970537262,
      "grad_norm": 6.956459874417088,
      "learning_rate": 4.923499528981739e-06,
      "loss": 0.4094,
      "step": 265
    },
    {
      "epoch": 0.4610051993067591,
      "grad_norm": 7.199980142354073,
      "learning_rate": 4.926823036901976e-06,
      "loss": 0.5051,
      "step": 266
    },
    {
      "epoch": 0.462738301559792,
      "grad_norm": 12.184218683876429,
      "learning_rate": 4.930134073843848e-06,
      "loss": 0.4363,
      "step": 267
    },
    {
      "epoch": 0.464471403812825,
      "grad_norm": 11.637578906192282,
      "learning_rate": 4.933432733048643e-06,
      "loss": 0.4845,
      "step": 268
    },
    {
      "epoch": 0.4662045060658579,
      "grad_norm": 11.991368050178584,
      "learning_rate": 4.9367191067158396e-06,
      "loss": 0.527,
      "step": 269
    },
    {
      "epoch": 0.4679376083188908,
      "grad_norm": 14.921181675646746,
      "learning_rate": 4.9399932860185734e-06,
      "loss": 0.4879,
      "step": 270
    },
    {
      "epoch": 0.4696707105719237,
      "grad_norm": 10.781114483410077,
      "learning_rate": 4.943255361118814e-06,
      "loss": 0.4241,
      "step": 271
    },
    {
      "epoch": 0.4714038128249567,
      "grad_norm": 7.495748713724125,
      "learning_rate": 4.946505421182261e-06,
      "loss": 0.4591,
      "step": 272
    },
    {
      "epoch": 0.4731369150779896,
      "grad_norm": 13.744091637195986,
      "learning_rate": 4.949743554392967e-06,
      "loss": 0.4966,
      "step": 273
    },
    {
      "epoch": 0.4748700173310225,
      "grad_norm": 6.999905449364929,
      "learning_rate": 4.952969847967701e-06,
      "loss": 0.3592,
      "step": 274
    },
    {
      "epoch": 0.47660311958405543,
      "grad_norm": 6.4023806912448595,
      "learning_rate": 4.956184388170033e-06,
      "loss": 0.3913,
      "step": 275
    },
    {
      "epoch": 0.4783362218370884,
      "grad_norm": 7.436839098376843,
      "learning_rate": 4.9593872603241835e-06,
      "loss": 0.3999,
      "step": 276
    },
    {
      "epoch": 0.4800693240901213,
      "grad_norm": 12.053729253472373,
      "learning_rate": 4.9625785488286085e-06,
      "loss": 0.4177,
      "step": 277
    },
    {
      "epoch": 0.48180242634315423,
      "grad_norm": 14.189186836154109,
      "learning_rate": 4.965758337169345e-06,
      "loss": 0.3504,
      "step": 278
    },
    {
      "epoch": 0.48353552859618715,
      "grad_norm": 4.932507272242248,
      "learning_rate": 4.968926707933111e-06,
      "loss": 0.3311,
      "step": 279
    },
    {
      "epoch": 0.4852686308492201,
      "grad_norm": 9.119063863392388,
      "learning_rate": 4.972083742820185e-06,
      "loss": 0.3483,
      "step": 280
    },
    {
      "epoch": 0.48700173310225303,
      "grad_norm": 11.13507172799957,
      "learning_rate": 4.9752295226570405e-06,
      "loss": 0.3919,
      "step": 281
    },
    {
      "epoch": 0.48873483535528595,
      "grad_norm": 12.633966566603487,
      "learning_rate": 4.9783641274087625e-06,
      "loss": 0.4415,
      "step": 282
    },
    {
      "epoch": 0.4904679376083189,
      "grad_norm": 12.113303294280167,
      "learning_rate": 4.981487636191247e-06,
      "loss": 0.4445,
      "step": 283
    },
    {
      "epoch": 0.49220103986135183,
      "grad_norm": 10.419482000731813,
      "learning_rate": 4.984600127283179e-06,
      "loss": 0.477,
      "step": 284
    },
    {
      "epoch": 0.49393414211438474,
      "grad_norm": 8.85462525619997,
      "learning_rate": 4.987701678137811e-06,
      "loss": 0.402,
      "step": 285
    },
    {
      "epoch": 0.49566724436741766,
      "grad_norm": 8.381579647081516,
      "learning_rate": 4.9907923653945165e-06,
      "loss": 0.4651,
      "step": 286
    },
    {
      "epoch": 0.49740034662045063,
      "grad_norm": 17.055922138086487,
      "learning_rate": 4.993872264890165e-06,
      "loss": 0.6026,
      "step": 287
    },
    {
      "epoch": 0.49913344887348354,
      "grad_norm": 8.918591965877159,
      "learning_rate": 4.996941451670277e-06,
      "loss": 0.4171,
      "step": 288
    },
    {
      "epoch": 0.5008665511265165,
      "grad_norm": 41.4727899901889,
      "learning_rate": 5e-06,
      "loss": 0.4336,
      "step": 289
    },
    {
      "epoch": 0.5025996533795494,
      "grad_norm": 10.994729253151151,
      "learning_rate": 5e-06,
      "loss": 0.3652,
      "step": 290
    },
    {
      "epoch": 0.5043327556325823,
      "grad_norm": 9.228475735694214,
      "learning_rate": 5e-06,
      "loss": 0.4648,
      "step": 291
    },
    {
      "epoch": 0.5060658578856152,
      "grad_norm": 5.17878929393979,
      "learning_rate": 5e-06,
      "loss": 0.3864,
      "step": 292
    },
    {
      "epoch": 0.5077989601386482,
      "grad_norm": 11.475091911396614,
      "learning_rate": 5e-06,
      "loss": 0.4303,
      "step": 293
    },
    {
      "epoch": 0.5095320623916811,
      "grad_norm": 13.317709606390892,
      "learning_rate": 5e-06,
      "loss": 0.4515,
      "step": 294
    },
    {
      "epoch": 0.511265164644714,
      "grad_norm": 11.481186446178551,
      "learning_rate": 5e-06,
      "loss": 0.5606,
      "step": 295
    },
    {
      "epoch": 0.512998266897747,
      "grad_norm": 5.439478196474985,
      "learning_rate": 5e-06,
      "loss": 0.417,
      "step": 296
    },
    {
      "epoch": 0.5147313691507799,
      "grad_norm": 6.794168990339224,
      "learning_rate": 5e-06,
      "loss": 0.5234,
      "step": 297
    },
    {
      "epoch": 0.5164644714038128,
      "grad_norm": 11.383421746092607,
      "learning_rate": 5e-06,
      "loss": 0.3842,
      "step": 298
    },
    {
      "epoch": 0.5181975736568457,
      "grad_norm": 6.089503290374867,
      "learning_rate": 5e-06,
      "loss": 0.3907,
      "step": 299
    },
    {
      "epoch": 0.5199306759098787,
      "grad_norm": 18.355479972464117,
      "learning_rate": 5e-06,
      "loss": 0.4565,
      "step": 300
    },
    {
      "epoch": 0.5216637781629117,
      "grad_norm": 13.156385697414796,
      "learning_rate": 5e-06,
      "loss": 0.6013,
      "step": 301
    },
    {
      "epoch": 0.5233968804159446,
      "grad_norm": 6.7702612375994615,
      "learning_rate": 5e-06,
      "loss": 0.3648,
      "step": 302
    },
    {
      "epoch": 0.5251299826689775,
      "grad_norm": 15.40924724030007,
      "learning_rate": 5e-06,
      "loss": 0.3752,
      "step": 303
    },
    {
      "epoch": 0.5268630849220104,
      "grad_norm": 8.673966439099306,
      "learning_rate": 5e-06,
      "loss": 0.3912,
      "step": 304
    },
    {
      "epoch": 0.5285961871750433,
      "grad_norm": 13.228721374867156,
      "learning_rate": 5e-06,
      "loss": 0.4171,
      "step": 305
    },
    {
      "epoch": 0.5303292894280762,
      "grad_norm": 8.364434253192297,
      "learning_rate": 5e-06,
      "loss": 0.4243,
      "step": 306
    },
    {
      "epoch": 0.5320623916811091,
      "grad_norm": 5.681877626975076,
      "learning_rate": 5e-06,
      "loss": 0.4518,
      "step": 307
    },
    {
      "epoch": 0.5337954939341422,
      "grad_norm": 12.848393772537472,
      "learning_rate": 5e-06,
      "loss": 0.5569,
      "step": 308
    },
    {
      "epoch": 0.5355285961871751,
      "grad_norm": 6.189584400232977,
      "learning_rate": 5e-06,
      "loss": 0.452,
      "step": 309
    },
    {
      "epoch": 0.537261698440208,
      "grad_norm": 6.2915655805875526,
      "learning_rate": 5e-06,
      "loss": 0.424,
      "step": 310
    },
    {
      "epoch": 0.5389948006932409,
      "grad_norm": 10.85608018249271,
      "learning_rate": 5e-06,
      "loss": 0.3745,
      "step": 311
    },
    {
      "epoch": 0.5407279029462738,
      "grad_norm": 9.933630615838,
      "learning_rate": 5e-06,
      "loss": 0.4864,
      "step": 312
    },
    {
      "epoch": 0.5424610051993067,
      "grad_norm": 11.693127587950366,
      "learning_rate": 5e-06,
      "loss": 0.4474,
      "step": 313
    },
    {
      "epoch": 0.5441941074523396,
      "grad_norm": 6.314642816320261,
      "learning_rate": 5e-06,
      "loss": 0.4444,
      "step": 314
    },
    {
      "epoch": 0.5459272097053726,
      "grad_norm": 8.059318920353068,
      "learning_rate": 5e-06,
      "loss": 0.4729,
      "step": 315
    },
    {
      "epoch": 0.5476603119584056,
      "grad_norm": 13.224364042224831,
      "learning_rate": 5e-06,
      "loss": 0.4198,
      "step": 316
    },
    {
      "epoch": 0.5493934142114385,
      "grad_norm": 10.847207169651904,
      "learning_rate": 5e-06,
      "loss": 0.4187,
      "step": 317
    },
    {
      "epoch": 0.5511265164644714,
      "grad_norm": 7.216102015347236,
      "learning_rate": 5e-06,
      "loss": 0.3257,
      "step": 318
    },
    {
      "epoch": 0.5528596187175043,
      "grad_norm": 10.448607604791475,
      "learning_rate": 5e-06,
      "loss": 0.4127,
      "step": 319
    },
    {
      "epoch": 0.5545927209705372,
      "grad_norm": 5.228076390543785,
      "learning_rate": 5e-06,
      "loss": 0.3971,
      "step": 320
    },
    {
      "epoch": 0.5563258232235702,
      "grad_norm": 15.234756105008756,
      "learning_rate": 5e-06,
      "loss": 0.5498,
      "step": 321
    },
    {
      "epoch": 0.5580589254766031,
      "grad_norm": 7.9468021703001215,
      "learning_rate": 5e-06,
      "loss": 0.395,
      "step": 322
    },
    {
      "epoch": 0.5597920277296361,
      "grad_norm": 59.53882005674953,
      "learning_rate": 5e-06,
      "loss": 0.5788,
      "step": 323
    },
    {
      "epoch": 0.561525129982669,
      "grad_norm": 8.295372778464595,
      "learning_rate": 5e-06,
      "loss": 0.5409,
      "step": 324
    },
    {
      "epoch": 0.5632582322357019,
      "grad_norm": 6.677292779344334,
      "learning_rate": 5e-06,
      "loss": 0.3252,
      "step": 325
    },
    {
      "epoch": 0.5649913344887348,
      "grad_norm": 13.439571629009947,
      "learning_rate": 5e-06,
      "loss": 0.5649,
      "step": 326
    },
    {
      "epoch": 0.5667244367417678,
      "grad_norm": 8.379746643460871,
      "learning_rate": 5e-06,
      "loss": 0.4868,
      "step": 327
    },
    {
      "epoch": 0.5684575389948007,
      "grad_norm": 8.587399109271331,
      "learning_rate": 5e-06,
      "loss": 0.3547,
      "step": 328
    },
    {
      "epoch": 0.5701906412478336,
      "grad_norm": 6.905244641897427,
      "learning_rate": 5e-06,
      "loss": 0.4326,
      "step": 329
    },
    {
      "epoch": 0.5719237435008665,
      "grad_norm": 10.513871974979056,
      "learning_rate": 5e-06,
      "loss": 0.5385,
      "step": 330
    },
    {
      "epoch": 0.5736568457538995,
      "grad_norm": 8.384763122743045,
      "learning_rate": 5e-06,
      "loss": 0.3926,
      "step": 331
    },
    {
      "epoch": 0.5753899480069324,
      "grad_norm": 11.09987692549924,
      "learning_rate": 5e-06,
      "loss": 0.4351,
      "step": 332
    },
    {
      "epoch": 0.5771230502599654,
      "grad_norm": 12.913111896917206,
      "learning_rate": 5e-06,
      "loss": 0.3973,
      "step": 333
    },
    {
      "epoch": 0.5788561525129983,
      "grad_norm": 15.149772093415196,
      "learning_rate": 5e-06,
      "loss": 0.6371,
      "step": 334
    },
    {
      "epoch": 0.5805892547660312,
      "grad_norm": 9.53337418696402,
      "learning_rate": 5e-06,
      "loss": 0.4306,
      "step": 335
    },
    {
      "epoch": 0.5823223570190641,
      "grad_norm": 16.011810711952307,
      "learning_rate": 5e-06,
      "loss": 0.5146,
      "step": 336
    },
    {
      "epoch": 0.584055459272097,
      "grad_norm": 15.860370292814583,
      "learning_rate": 5e-06,
      "loss": 0.5001,
      "step": 337
    },
    {
      "epoch": 0.58578856152513,
      "grad_norm": 6.952853148214391,
      "learning_rate": 5e-06,
      "loss": 0.3834,
      "step": 338
    },
    {
      "epoch": 0.587521663778163,
      "grad_norm": 6.942519562871924,
      "learning_rate": 5e-06,
      "loss": 0.366,
      "step": 339
    },
    {
      "epoch": 0.5892547660311959,
      "grad_norm": 6.883547099614226,
      "learning_rate": 5e-06,
      "loss": 0.4708,
      "step": 340
    },
    {
      "epoch": 0.5909878682842288,
      "grad_norm": 11.935243722882637,
      "learning_rate": 5e-06,
      "loss": 0.4065,
      "step": 341
    },
    {
      "epoch": 0.5927209705372617,
      "grad_norm": 9.564139032454982,
      "learning_rate": 5e-06,
      "loss": 0.3463,
      "step": 342
    },
    {
      "epoch": 0.5944540727902946,
      "grad_norm": 7.997880893424232,
      "learning_rate": 5e-06,
      "loss": 0.4119,
      "step": 343
    },
    {
      "epoch": 0.5961871750433275,
      "grad_norm": 6.7643689405817184,
      "learning_rate": 5e-06,
      "loss": 0.3837,
      "step": 344
    },
    {
      "epoch": 0.5979202772963604,
      "grad_norm": 8.846429533474092,
      "learning_rate": 5e-06,
      "loss": 0.4211,
      "step": 345
    },
    {
      "epoch": 0.5996533795493935,
      "grad_norm": 9.807183836217371,
      "learning_rate": 5e-06,
      "loss": 0.3394,
      "step": 346
    },
    {
      "epoch": 0.6013864818024264,
      "grad_norm": 18.704827584107047,
      "learning_rate": 5e-06,
      "loss": 0.4987,
      "step": 347
    },
    {
      "epoch": 0.6031195840554593,
      "grad_norm": 8.0869691936934,
      "learning_rate": 5e-06,
      "loss": 0.4209,
      "step": 348
    },
    {
      "epoch": 0.6048526863084922,
      "grad_norm": 7.2479263497120066,
      "learning_rate": 5e-06,
      "loss": 0.2938,
      "step": 349
    },
    {
      "epoch": 0.6065857885615251,
      "grad_norm": 10.718160273841855,
      "learning_rate": 5e-06,
      "loss": 0.476,
      "step": 350
    },
    {
      "epoch": 0.608318890814558,
      "grad_norm": 8.08732343949709,
      "learning_rate": 5e-06,
      "loss": 0.369,
      "step": 351
    },
    {
      "epoch": 0.610051993067591,
      "grad_norm": 10.982520347070983,
      "learning_rate": 5e-06,
      "loss": 0.5008,
      "step": 352
    },
    {
      "epoch": 0.6117850953206239,
      "grad_norm": 13.643108128797577,
      "learning_rate": 5e-06,
      "loss": 0.5958,
      "step": 353
    },
    {
      "epoch": 0.6135181975736569,
      "grad_norm": 11.465394072183859,
      "learning_rate": 5e-06,
      "loss": 0.4447,
      "step": 354
    },
    {
      "epoch": 0.6152512998266898,
      "grad_norm": 4.934536781098967,
      "learning_rate": 5e-06,
      "loss": 0.4335,
      "step": 355
    },
    {
      "epoch": 0.6169844020797227,
      "grad_norm": 9.314122410712578,
      "learning_rate": 5e-06,
      "loss": 0.2788,
      "step": 356
    },
    {
      "epoch": 0.6187175043327556,
      "grad_norm": 15.363397950389466,
      "learning_rate": 5e-06,
      "loss": 0.422,
      "step": 357
    },
    {
      "epoch": 0.6204506065857885,
      "grad_norm": 14.814076013863682,
      "learning_rate": 5e-06,
      "loss": 0.576,
      "step": 358
    },
    {
      "epoch": 0.6221837088388215,
      "grad_norm": 7.078451052265348,
      "learning_rate": 5e-06,
      "loss": 0.4,
      "step": 359
    },
    {
      "epoch": 0.6239168110918544,
      "grad_norm": 4.726705955661319,
      "learning_rate": 5e-06,
      "loss": 0.4163,
      "step": 360
    },
    {
      "epoch": 0.6256499133448874,
      "grad_norm": 5.5216137412440505,
      "learning_rate": 5e-06,
      "loss": 0.5234,
      "step": 361
    },
    {
      "epoch": 0.6273830155979203,
      "grad_norm": 9.833991939234926,
      "learning_rate": 5e-06,
      "loss": 0.4284,
      "step": 362
    },
    {
      "epoch": 0.6291161178509532,
      "grad_norm": 9.047141906279755,
      "learning_rate": 5e-06,
      "loss": 0.5817,
      "step": 363
    },
    {
      "epoch": 0.6308492201039861,
      "grad_norm": 4.582251033731436,
      "learning_rate": 5e-06,
      "loss": 0.3973,
      "step": 364
    },
    {
      "epoch": 0.6325823223570191,
      "grad_norm": 5.414578940240659,
      "learning_rate": 5e-06,
      "loss": 0.3645,
      "step": 365
    },
    {
      "epoch": 0.634315424610052,
      "grad_norm": 8.102698133400727,
      "learning_rate": 5e-06,
      "loss": 0.4423,
      "step": 366
    },
    {
      "epoch": 0.6360485268630849,
      "grad_norm": 7.356470138121027,
      "learning_rate": 5e-06,
      "loss": 0.3454,
      "step": 367
    },
    {
      "epoch": 0.6377816291161178,
      "grad_norm": 9.83435656724917,
      "learning_rate": 5e-06,
      "loss": 0.3035,
      "step": 368
    },
    {
      "epoch": 0.6395147313691508,
      "grad_norm": 11.581738112858698,
      "learning_rate": 5e-06,
      "loss": 0.5962,
      "step": 369
    },
    {
      "epoch": 0.6412478336221837,
      "grad_norm": 8.894548834111955,
      "learning_rate": 5e-06,
      "loss": 0.374,
      "step": 370
    },
    {
      "epoch": 0.6429809358752167,
      "grad_norm": 34.515904395655056,
      "learning_rate": 5e-06,
      "loss": 0.4113,
      "step": 371
    },
    {
      "epoch": 0.6447140381282496,
      "grad_norm": 5.731288484633433,
      "learning_rate": 5e-06,
      "loss": 0.374,
      "step": 372
    },
    {
      "epoch": 0.6464471403812825,
      "grad_norm": 9.373423939787449,
      "learning_rate": 5e-06,
      "loss": 0.3663,
      "step": 373
    },
    {
      "epoch": 0.6481802426343154,
      "grad_norm": 6.217401252824451,
      "learning_rate": 5e-06,
      "loss": 0.3779,
      "step": 374
    },
    {
      "epoch": 0.6499133448873483,
      "grad_norm": 12.691229511696442,
      "learning_rate": 5e-06,
      "loss": 0.4831,
      "step": 375
    },
    {
      "epoch": 0.6516464471403813,
      "grad_norm": 12.68058761438498,
      "learning_rate": 5e-06,
      "loss": 0.3083,
      "step": 376
    },
    {
      "epoch": 0.6533795493934142,
      "grad_norm": 9.287409013633644,
      "learning_rate": 5e-06,
      "loss": 0.4391,
      "step": 377
    },
    {
      "epoch": 0.6551126516464472,
      "grad_norm": 14.763353413229199,
      "learning_rate": 5e-06,
      "loss": 0.4219,
      "step": 378
    },
    {
      "epoch": 0.6568457538994801,
      "grad_norm": 4.473319751550226,
      "learning_rate": 5e-06,
      "loss": 0.366,
      "step": 379
    },
    {
      "epoch": 0.658578856152513,
      "grad_norm": 7.353392159488076,
      "learning_rate": 5e-06,
      "loss": 0.4755,
      "step": 380
    },
    {
      "epoch": 0.6603119584055459,
      "grad_norm": 14.515448397741627,
      "learning_rate": 5e-06,
      "loss": 0.4145,
      "step": 381
    },
    {
      "epoch": 0.6620450606585788,
      "grad_norm": 17.292670947712484,
      "learning_rate": 5e-06,
      "loss": 0.5588,
      "step": 382
    },
    {
      "epoch": 0.6637781629116117,
      "grad_norm": 8.810928962217666,
      "learning_rate": 5e-06,
      "loss": 0.3421,
      "step": 383
    },
    {
      "epoch": 0.6655112651646448,
      "grad_norm": 8.491463357465065,
      "learning_rate": 5e-06,
      "loss": 0.3541,
      "step": 384
    },
    {
      "epoch": 0.6672443674176777,
      "grad_norm": 11.800426953883715,
      "learning_rate": 5e-06,
      "loss": 0.4887,
      "step": 385
    },
    {
      "epoch": 0.6689774696707106,
      "grad_norm": 7.619308849238748,
      "learning_rate": 5e-06,
      "loss": 0.436,
      "step": 386
    },
    {
      "epoch": 0.6707105719237435,
      "grad_norm": 11.106509841172013,
      "learning_rate": 5e-06,
      "loss": 0.62,
      "step": 387
    },
    {
      "epoch": 0.6724436741767764,
      "grad_norm": 4.481657367742521,
      "learning_rate": 5e-06,
      "loss": 0.4077,
      "step": 388
    },
    {
      "epoch": 0.6741767764298093,
      "grad_norm": 10.611496905580685,
      "learning_rate": 5e-06,
      "loss": 0.4775,
      "step": 389
    },
    {
      "epoch": 0.6759098786828422,
      "grad_norm": 11.177595645270932,
      "learning_rate": 5e-06,
      "loss": 0.4691,
      "step": 390
    },
    {
      "epoch": 0.6776429809358753,
      "grad_norm": 7.64488379241064,
      "learning_rate": 5e-06,
      "loss": 0.5323,
      "step": 391
    },
    {
      "epoch": 0.6793760831889082,
      "grad_norm": 9.761950089795791,
      "learning_rate": 5e-06,
      "loss": 0.433,
      "step": 392
    },
    {
      "epoch": 0.6811091854419411,
      "grad_norm": 7.51069108000506,
      "learning_rate": 5e-06,
      "loss": 0.3459,
      "step": 393
    },
    {
      "epoch": 0.682842287694974,
      "grad_norm": 9.270401596703357,
      "learning_rate": 5e-06,
      "loss": 0.4645,
      "step": 394
    },
    {
      "epoch": 0.6845753899480069,
      "grad_norm": 8.412621108114367,
      "learning_rate": 5e-06,
      "loss": 0.3391,
      "step": 395
    },
    {
      "epoch": 0.6863084922010398,
      "grad_norm": 12.358757071112509,
      "learning_rate": 5e-06,
      "loss": 0.4453,
      "step": 396
    },
    {
      "epoch": 0.6880415944540728,
      "grad_norm": 10.203193792838531,
      "learning_rate": 5e-06,
      "loss": 0.3989,
      "step": 397
    },
    {
      "epoch": 0.6897746967071057,
      "grad_norm": 7.438457010710979,
      "learning_rate": 5e-06,
      "loss": 0.4196,
      "step": 398
    },
    {
      "epoch": 0.6915077989601387,
      "grad_norm": 13.12057158830022,
      "learning_rate": 5e-06,
      "loss": 0.46,
      "step": 399
    },
    {
      "epoch": 0.6932409012131716,
      "grad_norm": 14.11149613329916,
      "learning_rate": 5e-06,
      "loss": 0.4607,
      "step": 400
    },
    {
      "epoch": 0.6949740034662045,
      "grad_norm": 6.348597886887665,
      "learning_rate": 5e-06,
      "loss": 0.3694,
      "step": 401
    },
    {
      "epoch": 0.6967071057192374,
      "grad_norm": 9.695174001831079,
      "learning_rate": 5e-06,
      "loss": 0.4265,
      "step": 402
    },
    {
      "epoch": 0.6984402079722704,
      "grad_norm": 10.28456169058655,
      "learning_rate": 5e-06,
      "loss": 0.4112,
      "step": 403
    },
    {
      "epoch": 0.7001733102253033,
      "grad_norm": 12.136485529421579,
      "learning_rate": 5e-06,
      "loss": 0.3948,
      "step": 404
    },
    {
      "epoch": 0.7019064124783362,
      "grad_norm": 4.9227772687993925,
      "learning_rate": 5e-06,
      "loss": 0.4279,
      "step": 405
    },
    {
      "epoch": 0.7036395147313691,
      "grad_norm": 7.644299955382529,
      "learning_rate": 5e-06,
      "loss": 0.4587,
      "step": 406
    },
    {
      "epoch": 0.7053726169844021,
      "grad_norm": 7.048433321843973,
      "learning_rate": 5e-06,
      "loss": 0.3931,
      "step": 407
    },
    {
      "epoch": 0.707105719237435,
      "grad_norm": 9.25989683485134,
      "learning_rate": 5e-06,
      "loss": 0.4381,
      "step": 408
    },
    {
      "epoch": 0.708838821490468,
      "grad_norm": 9.6827506944433,
      "learning_rate": 5e-06,
      "loss": 0.4645,
      "step": 409
    },
    {
      "epoch": 0.7105719237435009,
      "grad_norm": 6.6684081028365805,
      "learning_rate": 5e-06,
      "loss": 0.3705,
      "step": 410
    },
    {
      "epoch": 0.7123050259965338,
      "grad_norm": 5.9324978286604155,
      "learning_rate": 5e-06,
      "loss": 0.3587,
      "step": 411
    },
    {
      "epoch": 0.7140381282495667,
      "grad_norm": 7.704414925117864,
      "learning_rate": 5e-06,
      "loss": 0.5886,
      "step": 412
    },
    {
      "epoch": 0.7157712305025996,
      "grad_norm": 6.074079015173276,
      "learning_rate": 5e-06,
      "loss": 0.3726,
      "step": 413
    },
    {
      "epoch": 0.7175043327556326,
      "grad_norm": 9.153251078718348,
      "learning_rate": 5e-06,
      "loss": 0.3529,
      "step": 414
    },
    {
      "epoch": 0.7192374350086655,
      "grad_norm": 10.851840911655266,
      "learning_rate": 5e-06,
      "loss": 0.5035,
      "step": 415
    },
    {
      "epoch": 0.7209705372616985,
      "grad_norm": 10.680119336208024,
      "learning_rate": 5e-06,
      "loss": 0.3844,
      "step": 416
    },
    {
      "epoch": 0.7227036395147314,
      "grad_norm": 9.627496568970923,
      "learning_rate": 5e-06,
      "loss": 0.4723,
      "step": 417
    },
    {
      "epoch": 0.7244367417677643,
      "grad_norm": 6.1864660150501996,
      "learning_rate": 5e-06,
      "loss": 0.365,
      "step": 418
    },
    {
      "epoch": 0.7261698440207972,
      "grad_norm": 7.782822783210055,
      "learning_rate": 5e-06,
      "loss": 0.3692,
      "step": 419
    },
    {
      "epoch": 0.7279029462738301,
      "grad_norm": 8.214334324284183,
      "learning_rate": 5e-06,
      "loss": 0.3964,
      "step": 420
    },
    {
      "epoch": 0.729636048526863,
      "grad_norm": 7.997683666585077,
      "learning_rate": 5e-06,
      "loss": 0.3289,
      "step": 421
    },
    {
      "epoch": 0.7313691507798961,
      "grad_norm": 9.60161653259817,
      "learning_rate": 5e-06,
      "loss": 0.4668,
      "step": 422
    },
    {
      "epoch": 0.733102253032929,
      "grad_norm": 6.984824488297771,
      "learning_rate": 5e-06,
      "loss": 0.4453,
      "step": 423
    },
    {
      "epoch": 0.7348353552859619,
      "grad_norm": 9.18436591221586,
      "learning_rate": 5e-06,
      "loss": 0.445,
      "step": 424
    },
    {
      "epoch": 0.7365684575389948,
      "grad_norm": 12.113696935386109,
      "learning_rate": 5e-06,
      "loss": 0.5339,
      "step": 425
    },
    {
      "epoch": 0.7383015597920277,
      "grad_norm": 5.131921142103087,
      "learning_rate": 5e-06,
      "loss": 0.4295,
      "step": 426
    },
    {
      "epoch": 0.7400346620450606,
      "grad_norm": 8.236802150590286,
      "learning_rate": 5e-06,
      "loss": 0.4105,
      "step": 427
    },
    {
      "epoch": 0.7417677642980935,
      "grad_norm": 9.146593437383618,
      "learning_rate": 5e-06,
      "loss": 0.4973,
      "step": 428
    },
    {
      "epoch": 0.7435008665511266,
      "grad_norm": 4.928455422499274,
      "learning_rate": 5e-06,
      "loss": 0.4908,
      "step": 429
    },
    {
      "epoch": 0.7452339688041595,
      "grad_norm": 5.5499367443982,
      "learning_rate": 5e-06,
      "loss": 0.4054,
      "step": 430
    },
    {
      "epoch": 0.7469670710571924,
      "grad_norm": 8.028401029065257,
      "learning_rate": 5e-06,
      "loss": 0.3485,
      "step": 431
    },
    {
      "epoch": 0.7487001733102253,
      "grad_norm": 7.262274513092687,
      "learning_rate": 5e-06,
      "loss": 0.4179,
      "step": 432
    },
    {
      "epoch": 0.7504332755632582,
      "grad_norm": 9.091266306014525,
      "learning_rate": 5e-06,
      "loss": 0.4983,
      "step": 433
    },
    {
      "epoch": 0.7521663778162911,
      "grad_norm": 8.269364720660274,
      "learning_rate": 5e-06,
      "loss": 0.3846,
      "step": 434
    },
    {
      "epoch": 0.7538994800693241,
      "grad_norm": 9.966752668369042,
      "learning_rate": 5e-06,
      "loss": 0.3624,
      "step": 435
    },
    {
      "epoch": 0.755632582322357,
      "grad_norm": 7.725068629679089,
      "learning_rate": 5e-06,
      "loss": 0.3025,
      "step": 436
    },
    {
      "epoch": 0.75736568457539,
      "grad_norm": 11.985987429215847,
      "learning_rate": 5e-06,
      "loss": 0.3084,
      "step": 437
    },
    {
      "epoch": 0.7590987868284229,
      "grad_norm": 11.227313415431151,
      "learning_rate": 5e-06,
      "loss": 0.3608,
      "step": 438
    },
    {
      "epoch": 0.7608318890814558,
      "grad_norm": 5.058929691427104,
      "learning_rate": 5e-06,
      "loss": 0.2923,
      "step": 439
    },
    {
      "epoch": 0.7625649913344887,
      "grad_norm": 9.407390335258864,
      "learning_rate": 5e-06,
      "loss": 0.4917,
      "step": 440
    },
    {
      "epoch": 0.7642980935875217,
      "grad_norm": 5.886495179687849,
      "learning_rate": 5e-06,
      "loss": 0.2978,
      "step": 441
    },
    {
      "epoch": 0.7660311958405546,
      "grad_norm": 9.361742742291712,
      "learning_rate": 5e-06,
      "loss": 0.4797,
      "step": 442
    },
    {
      "epoch": 0.7677642980935875,
      "grad_norm": 6.556695605500044,
      "learning_rate": 5e-06,
      "loss": 0.2622,
      "step": 443
    },
    {
      "epoch": 0.7694974003466204,
      "grad_norm": 4.3856333530319835,
      "learning_rate": 5e-06,
      "loss": 0.2692,
      "step": 444
    },
    {
      "epoch": 0.7712305025996534,
      "grad_norm": 11.3204668349124,
      "learning_rate": 5e-06,
      "loss": 0.5927,
      "step": 445
    },
    {
      "epoch": 0.7729636048526863,
      "grad_norm": 7.312701100249606,
      "learning_rate": 5e-06,
      "loss": 0.3175,
      "step": 446
    },
    {
      "epoch": 0.7746967071057193,
      "grad_norm": 8.118712354518427,
      "learning_rate": 5e-06,
      "loss": 0.3607,
      "step": 447
    },
    {
      "epoch": 0.7764298093587522,
      "grad_norm": 14.592130844834603,
      "learning_rate": 5e-06,
      "loss": 0.4445,
      "step": 448
    },
    {
      "epoch": 0.7781629116117851,
      "grad_norm": 9.619337125213756,
      "learning_rate": 5e-06,
      "loss": 0.386,
      "step": 449
    },
    {
      "epoch": 0.779896013864818,
      "grad_norm": 8.988173449755552,
      "learning_rate": 5e-06,
      "loss": 0.4194,
      "step": 450
    },
    {
      "epoch": 0.7816291161178509,
      "grad_norm": 14.61401058641747,
      "learning_rate": 5e-06,
      "loss": 0.5514,
      "step": 451
    },
    {
      "epoch": 0.7833622183708839,
      "grad_norm": 16.336369031244367,
      "learning_rate": 5e-06,
      "loss": 0.5779,
      "step": 452
    },
    {
      "epoch": 0.7850953206239168,
      "grad_norm": 6.18960812808831,
      "learning_rate": 5e-06,
      "loss": 0.3657,
      "step": 453
    },
    {
      "epoch": 0.7868284228769498,
      "grad_norm": 10.143415117171228,
      "learning_rate": 5e-06,
      "loss": 0.4263,
      "step": 454
    },
    {
      "epoch": 0.7885615251299827,
      "grad_norm": 5.656875154290615,
      "learning_rate": 5e-06,
      "loss": 0.3117,
      "step": 455
    },
    {
      "epoch": 0.7902946273830156,
      "grad_norm": 9.957204416893564,
      "learning_rate": 5e-06,
      "loss": 0.3863,
      "step": 456
    },
    {
      "epoch": 0.7920277296360485,
      "grad_norm": 4.993022245963024,
      "learning_rate": 5e-06,
      "loss": 0.3009,
      "step": 457
    },
    {
      "epoch": 0.7937608318890814,
      "grad_norm": 9.938787412922052,
      "learning_rate": 5e-06,
      "loss": 0.4463,
      "step": 458
    },
    {
      "epoch": 0.7954939341421143,
      "grad_norm": 10.015407518236472,
      "learning_rate": 5e-06,
      "loss": 0.4345,
      "step": 459
    },
    {
      "epoch": 0.7972270363951474,
      "grad_norm": 6.09252103492759,
      "learning_rate": 5e-06,
      "loss": 0.4797,
      "step": 460
    },
    {
      "epoch": 0.7989601386481803,
      "grad_norm": 5.419340481204396,
      "learning_rate": 5e-06,
      "loss": 0.3687,
      "step": 461
    },
    {
      "epoch": 0.8006932409012132,
      "grad_norm": 12.445661311221386,
      "learning_rate": 5e-06,
      "loss": 0.4097,
      "step": 462
    },
    {
      "epoch": 0.8024263431542461,
      "grad_norm": 9.61355306465073,
      "learning_rate": 5e-06,
      "loss": 0.3964,
      "step": 463
    },
    {
      "epoch": 0.804159445407279,
      "grad_norm": 7.912640427824271,
      "learning_rate": 5e-06,
      "loss": 0.3289,
      "step": 464
    },
    {
      "epoch": 0.8058925476603119,
      "grad_norm": 7.722068907071887,
      "learning_rate": 5e-06,
      "loss": 0.315,
      "step": 465
    },
    {
      "epoch": 0.8076256499133448,
      "grad_norm": 8.13172000872466,
      "learning_rate": 5e-06,
      "loss": 0.4259,
      "step": 466
    },
    {
      "epoch": 0.8093587521663779,
      "grad_norm": 9.55431978480378,
      "learning_rate": 5e-06,
      "loss": 0.4101,
      "step": 467
    },
    {
      "epoch": 0.8110918544194108,
      "grad_norm": 7.403619355377893,
      "learning_rate": 5e-06,
      "loss": 0.4142,
      "step": 468
    },
    {
      "epoch": 0.8128249566724437,
      "grad_norm": 12.598531340730132,
      "learning_rate": 5e-06,
      "loss": 0.3685,
      "step": 469
    },
    {
      "epoch": 0.8145580589254766,
      "grad_norm": 6.33759328559071,
      "learning_rate": 5e-06,
      "loss": 0.3634,
      "step": 470
    },
    {
      "epoch": 0.8162911611785095,
      "grad_norm": 10.894037530270285,
      "learning_rate": 5e-06,
      "loss": 0.5497,
      "step": 471
    },
    {
      "epoch": 0.8180242634315424,
      "grad_norm": 13.819381173573143,
      "learning_rate": 5e-06,
      "loss": 0.41,
      "step": 472
    },
    {
      "epoch": 0.8197573656845754,
      "grad_norm": 7.161768755109324,
      "learning_rate": 5e-06,
      "loss": 0.4425,
      "step": 473
    },
    {
      "epoch": 0.8214904679376083,
      "grad_norm": 6.037285349566463,
      "learning_rate": 5e-06,
      "loss": 0.2261,
      "step": 474
    },
    {
      "epoch": 0.8232235701906413,
      "grad_norm": 8.839429467203416,
      "learning_rate": 5e-06,
      "loss": 0.5193,
      "step": 475
    },
    {
      "epoch": 0.8249566724436742,
      "grad_norm": 12.340079309836176,
      "learning_rate": 5e-06,
      "loss": 0.4481,
      "step": 476
    },
    {
      "epoch": 0.8266897746967071,
      "grad_norm": 9.006788448674682,
      "learning_rate": 5e-06,
      "loss": 0.2998,
      "step": 477
    },
    {
      "epoch": 0.82842287694974,
      "grad_norm": 9.429128002897167,
      "learning_rate": 5e-06,
      "loss": 0.4289,
      "step": 478
    },
    {
      "epoch": 0.830155979202773,
      "grad_norm": 7.506173708207996,
      "learning_rate": 5e-06,
      "loss": 0.5545,
      "step": 479
    },
    {
      "epoch": 0.8318890814558059,
      "grad_norm": 10.967257580205109,
      "learning_rate": 5e-06,
      "loss": 0.3928,
      "step": 480
    },
    {
      "epoch": 0.8336221837088388,
      "grad_norm": 8.016833714299828,
      "learning_rate": 5e-06,
      "loss": 0.3956,
      "step": 481
    },
    {
      "epoch": 0.8353552859618717,
      "grad_norm": 13.458403186597131,
      "learning_rate": 5e-06,
      "loss": 0.4853,
      "step": 482
    },
    {
      "epoch": 0.8370883882149047,
      "grad_norm": 6.9812815939547805,
      "learning_rate": 5e-06,
      "loss": 0.3584,
      "step": 483
    },
    {
      "epoch": 0.8388214904679376,
      "grad_norm": 9.574198031393331,
      "learning_rate": 5e-06,
      "loss": 0.3917,
      "step": 484
    },
    {
      "epoch": 0.8405545927209706,
      "grad_norm": 8.258029613690299,
      "learning_rate": 5e-06,
      "loss": 0.3983,
      "step": 485
    },
    {
      "epoch": 0.8422876949740035,
      "grad_norm": 4.206551102039801,
      "learning_rate": 5e-06,
      "loss": 0.352,
      "step": 486
    },
    {
      "epoch": 0.8440207972270364,
      "grad_norm": 14.727166039522473,
      "learning_rate": 5e-06,
      "loss": 0.4921,
      "step": 487
    },
    {
      "epoch": 0.8457538994800693,
      "grad_norm": 9.714256695295315,
      "learning_rate": 5e-06,
      "loss": 0.4595,
      "step": 488
    },
    {
      "epoch": 0.8474870017331022,
      "grad_norm": 8.397726623257324,
      "learning_rate": 5e-06,
      "loss": 0.326,
      "step": 489
    },
    {
      "epoch": 0.8492201039861352,
      "grad_norm": 8.067326010668117,
      "learning_rate": 5e-06,
      "loss": 0.3486,
      "step": 490
    },
    {
      "epoch": 0.8509532062391681,
      "grad_norm": 9.399968784361608,
      "learning_rate": 5e-06,
      "loss": 0.5605,
      "step": 491
    },
    {
      "epoch": 0.8526863084922011,
      "grad_norm": 12.862570679584415,
      "learning_rate": 5e-06,
      "loss": 0.5728,
      "step": 492
    },
    {
      "epoch": 0.854419410745234,
      "grad_norm": 3.844079228389517,
      "learning_rate": 5e-06,
      "loss": 0.2922,
      "step": 493
    },
    {
      "epoch": 0.8561525129982669,
      "grad_norm": 6.932927729895861,
      "learning_rate": 5e-06,
      "loss": 0.3511,
      "step": 494
    },
    {
      "epoch": 0.8578856152512998,
      "grad_norm": 5.479397419565674,
      "learning_rate": 5e-06,
      "loss": 0.4277,
      "step": 495
    },
    {
      "epoch": 0.8596187175043327,
      "grad_norm": 5.73243817874598,
      "learning_rate": 5e-06,
      "loss": 0.4421,
      "step": 496
    },
    {
      "epoch": 0.8613518197573656,
      "grad_norm": 9.477186416755222,
      "learning_rate": 5e-06,
      "loss": 0.2734,
      "step": 497
    },
    {
      "epoch": 0.8630849220103987,
      "grad_norm": 11.565832369248428,
      "learning_rate": 5e-06,
      "loss": 0.4028,
      "step": 498
    },
    {
      "epoch": 0.8648180242634316,
      "grad_norm": 9.915777684817593,
      "learning_rate": 5e-06,
      "loss": 0.306,
      "step": 499
    },
    {
      "epoch": 0.8665511265164645,
      "grad_norm": 8.293229561224974,
      "learning_rate": 5e-06,
      "loss": 0.427,
      "step": 500
    },
    {
      "epoch": 0.8682842287694974,
      "grad_norm": 7.637408306863786,
      "learning_rate": 5e-06,
      "loss": 0.4123,
      "step": 501
    },
    {
      "epoch": 0.8700173310225303,
      "grad_norm": 11.0278456840912,
      "learning_rate": 5e-06,
      "loss": 0.3991,
      "step": 502
    },
    {
      "epoch": 0.8717504332755632,
      "grad_norm": 10.421260054945972,
      "learning_rate": 5e-06,
      "loss": 0.3625,
      "step": 503
    },
    {
      "epoch": 0.8734835355285961,
      "grad_norm": 6.061881436911793,
      "learning_rate": 5e-06,
      "loss": 0.4495,
      "step": 504
    },
    {
      "epoch": 0.8752166377816292,
      "grad_norm": 10.451873932514298,
      "learning_rate": 5e-06,
      "loss": 0.43,
      "step": 505
    },
    {
      "epoch": 0.8769497400346621,
      "grad_norm": 7.868611241966225,
      "learning_rate": 5e-06,
      "loss": 0.3915,
      "step": 506
    },
    {
      "epoch": 0.878682842287695,
      "grad_norm": 6.370253347871217,
      "learning_rate": 5e-06,
      "loss": 0.3183,
      "step": 507
    },
    {
      "epoch": 0.8804159445407279,
      "grad_norm": 5.559812089228313,
      "learning_rate": 5e-06,
      "loss": 0.3541,
      "step": 508
    },
    {
      "epoch": 0.8821490467937608,
      "grad_norm": 5.9614385591754795,
      "learning_rate": 5e-06,
      "loss": 0.2478,
      "step": 509
    },
    {
      "epoch": 0.8838821490467937,
      "grad_norm": 7.793854458987598,
      "learning_rate": 5e-06,
      "loss": 0.3727,
      "step": 510
    },
    {
      "epoch": 0.8856152512998267,
      "grad_norm": 12.35560089243279,
      "learning_rate": 5e-06,
      "loss": 0.4056,
      "step": 511
    },
    {
      "epoch": 0.8873483535528596,
      "grad_norm": 13.132937401807894,
      "learning_rate": 5e-06,
      "loss": 0.5514,
      "step": 512
    },
    {
      "epoch": 0.8890814558058926,
      "grad_norm": 10.449983178814213,
      "learning_rate": 5e-06,
      "loss": 0.3636,
      "step": 513
    },
    {
      "epoch": 0.8908145580589255,
      "grad_norm": 4.470143987738892,
      "learning_rate": 5e-06,
      "loss": 0.4156,
      "step": 514
    },
    {
      "epoch": 0.8925476603119584,
      "grad_norm": 6.5463596040976935,
      "learning_rate": 5e-06,
      "loss": 0.3969,
      "step": 515
    },
    {
      "epoch": 0.8942807625649913,
      "grad_norm": 17.465087943523216,
      "learning_rate": 5e-06,
      "loss": 0.6034,
      "step": 516
    },
    {
      "epoch": 0.8960138648180243,
      "grad_norm": 10.222239420595306,
      "learning_rate": 5e-06,
      "loss": 0.3788,
      "step": 517
    },
    {
      "epoch": 0.8977469670710572,
      "grad_norm": 11.891604730292856,
      "learning_rate": 5e-06,
      "loss": 0.4512,
      "step": 518
    },
    {
      "epoch": 0.8994800693240901,
      "grad_norm": 11.406418556117593,
      "learning_rate": 5e-06,
      "loss": 0.4888,
      "step": 519
    },
    {
      "epoch": 0.901213171577123,
      "grad_norm": 8.688689884509614,
      "learning_rate": 5e-06,
      "loss": 0.4315,
      "step": 520
    },
    {
      "epoch": 0.902946273830156,
      "grad_norm": 5.113991427954813,
      "learning_rate": 5e-06,
      "loss": 0.4254,
      "step": 521
    },
    {
      "epoch": 0.9046793760831889,
      "grad_norm": 5.104572991008379,
      "learning_rate": 5e-06,
      "loss": 0.3417,
      "step": 522
    },
    {
      "epoch": 0.9064124783362218,
      "grad_norm": 7.149891256292409,
      "learning_rate": 5e-06,
      "loss": 0.4411,
      "step": 523
    },
    {
      "epoch": 0.9081455805892548,
      "grad_norm": 6.938182075581183,
      "learning_rate": 5e-06,
      "loss": 0.4139,
      "step": 524
    },
    {
      "epoch": 0.9098786828422877,
      "grad_norm": 14.837884326254093,
      "learning_rate": 5e-06,
      "loss": 0.5857,
      "step": 525
    },
    {
      "epoch": 0.9116117850953206,
      "grad_norm": 8.23761629013847,
      "learning_rate": 5e-06,
      "loss": 0.3928,
      "step": 526
    },
    {
      "epoch": 0.9133448873483535,
      "grad_norm": 10.54029815194219,
      "learning_rate": 5e-06,
      "loss": 0.408,
      "step": 527
    },
    {
      "epoch": 0.9150779896013865,
      "grad_norm": 7.106406864525132,
      "learning_rate": 5e-06,
      "loss": 0.4391,
      "step": 528
    },
    {
      "epoch": 0.9168110918544194,
      "grad_norm": 5.66805052221768,
      "learning_rate": 5e-06,
      "loss": 0.3597,
      "step": 529
    },
    {
      "epoch": 0.9185441941074524,
      "grad_norm": 6.779779371424657,
      "learning_rate": 5e-06,
      "loss": 0.3782,
      "step": 530
    },
    {
      "epoch": 0.9202772963604853,
      "grad_norm": 9.388700519706147,
      "learning_rate": 5e-06,
      "loss": 0.3168,
      "step": 531
    },
    {
      "epoch": 0.9220103986135182,
      "grad_norm": 11.211333902508441,
      "learning_rate": 5e-06,
      "loss": 0.4566,
      "step": 532
    },
    {
      "epoch": 0.9237435008665511,
      "grad_norm": 8.405341003508061,
      "learning_rate": 5e-06,
      "loss": 0.3798,
      "step": 533
    },
    {
      "epoch": 0.925476603119584,
      "grad_norm": 6.912815753147402,
      "learning_rate": 5e-06,
      "loss": 0.4281,
      "step": 534
    },
    {
      "epoch": 0.9272097053726169,
      "grad_norm": 8.392231537016745,
      "learning_rate": 5e-06,
      "loss": 0.5624,
      "step": 535
    },
    {
      "epoch": 0.92894280762565,
      "grad_norm": 5.9787309689451185,
      "learning_rate": 5e-06,
      "loss": 0.3868,
      "step": 536
    },
    {
      "epoch": 0.9306759098786829,
      "grad_norm": 7.446722643626759,
      "learning_rate": 5e-06,
      "loss": 0.3245,
      "step": 537
    },
    {
      "epoch": 0.9324090121317158,
      "grad_norm": 7.313524484190873,
      "learning_rate": 5e-06,
      "loss": 0.3418,
      "step": 538
    },
    {
      "epoch": 0.9341421143847487,
      "grad_norm": 5.634672029931085,
      "learning_rate": 5e-06,
      "loss": 0.3711,
      "step": 539
    },
    {
      "epoch": 0.9358752166377816,
      "grad_norm": 10.290873096671481,
      "learning_rate": 5e-06,
      "loss": 0.4392,
      "step": 540
    },
    {
      "epoch": 0.9376083188908145,
      "grad_norm": 9.985689987006484,
      "learning_rate": 5e-06,
      "loss": 0.427,
      "step": 541
    },
    {
      "epoch": 0.9393414211438474,
      "grad_norm": 5.088400436310427,
      "learning_rate": 5e-06,
      "loss": 0.4769,
      "step": 542
    },
    {
      "epoch": 0.9410745233968805,
      "grad_norm": 7.545706522265892,
      "learning_rate": 5e-06,
      "loss": 0.3318,
      "step": 543
    },
    {
      "epoch": 0.9428076256499134,
      "grad_norm": 4.859352430677593,
      "learning_rate": 5e-06,
      "loss": 0.4051,
      "step": 544
    },
    {
      "epoch": 0.9445407279029463,
      "grad_norm": 10.384084745581479,
      "learning_rate": 5e-06,
      "loss": 0.363,
      "step": 545
    },
    {
      "epoch": 0.9462738301559792,
      "grad_norm": 12.33963600846026,
      "learning_rate": 5e-06,
      "loss": 0.4387,
      "step": 546
    },
    {
      "epoch": 0.9480069324090121,
      "grad_norm": 8.623918106114909,
      "learning_rate": 5e-06,
      "loss": 0.3748,
      "step": 547
    },
    {
      "epoch": 0.949740034662045,
      "grad_norm": 6.358320968969213,
      "learning_rate": 5e-06,
      "loss": 0.3689,
      "step": 548
    },
    {
      "epoch": 0.951473136915078,
      "grad_norm": 6.292139132446178,
      "learning_rate": 5e-06,
      "loss": 0.379,
      "step": 549
    },
    {
      "epoch": 0.9532062391681109,
      "grad_norm": 6.406271734433405,
      "learning_rate": 5e-06,
      "loss": 0.3013,
      "step": 550
    },
    {
      "epoch": 0.9549393414211439,
      "grad_norm": 6.99165173401297,
      "learning_rate": 5e-06,
      "loss": 0.4482,
      "step": 551
    },
    {
      "epoch": 0.9566724436741768,
      "grad_norm": 12.408519008896292,
      "learning_rate": 5e-06,
      "loss": 0.4414,
      "step": 552
    },
    {
      "epoch": 0.9584055459272097,
      "grad_norm": 6.8532690447947395,
      "learning_rate": 5e-06,
      "loss": 0.4047,
      "step": 553
    },
    {
      "epoch": 0.9601386481802426,
      "grad_norm": 8.062491956603303,
      "learning_rate": 5e-06,
      "loss": 0.3473,
      "step": 554
    },
    {
      "epoch": 0.9618717504332756,
      "grad_norm": 7.701155528767841,
      "learning_rate": 5e-06,
      "loss": 0.3716,
      "step": 555
    },
    {
      "epoch": 0.9636048526863085,
      "grad_norm": 6.704961778501938,
      "learning_rate": 5e-06,
      "loss": 0.3161,
      "step": 556
    },
    {
      "epoch": 0.9653379549393414,
      "grad_norm": 8.665487160258756,
      "learning_rate": 5e-06,
      "loss": 0.3351,
      "step": 557
    },
    {
      "epoch": 0.9670710571923743,
      "grad_norm": 8.472376198402998,
      "learning_rate": 5e-06,
      "loss": 0.2966,
      "step": 558
    },
    {
      "epoch": 0.9688041594454073,
      "grad_norm": 7.772409096938093,
      "learning_rate": 5e-06,
      "loss": 0.3524,
      "step": 559
    },
    {
      "epoch": 0.9705372616984402,
      "grad_norm": 6.907995219252743,
      "learning_rate": 5e-06,
      "loss": 0.3456,
      "step": 560
    },
    {
      "epoch": 0.9722703639514731,
      "grad_norm": 6.366812441024134,
      "learning_rate": 5e-06,
      "loss": 0.3267,
      "step": 561
    },
    {
      "epoch": 0.9740034662045061,
      "grad_norm": 9.254428525125695,
      "learning_rate": 5e-06,
      "loss": 0.4101,
      "step": 562
    },
    {
      "epoch": 0.975736568457539,
      "grad_norm": 6.794071855966288,
      "learning_rate": 5e-06,
      "loss": 0.2768,
      "step": 563
    },
    {
      "epoch": 0.9774696707105719,
      "grad_norm": 10.365106852487173,
      "learning_rate": 5e-06,
      "loss": 0.468,
      "step": 564
    },
    {
      "epoch": 0.9792027729636048,
      "grad_norm": 16.61247343659699,
      "learning_rate": 5e-06,
      "loss": 0.5585,
      "step": 565
    },
    {
      "epoch": 0.9809358752166378,
      "grad_norm": 6.558142386940284,
      "learning_rate": 5e-06,
      "loss": 0.3861,
      "step": 566
    },
    {
      "epoch": 0.9826689774696707,
      "grad_norm": 8.723147797390412,
      "learning_rate": 5e-06,
      "loss": 0.4417,
      "step": 567
    },
    {
      "epoch": 0.9844020797227037,
      "grad_norm": 12.431657655379144,
      "learning_rate": 5e-06,
      "loss": 0.544,
      "step": 568
    },
    {
      "epoch": 0.9861351819757366,
      "grad_norm": 8.122719782200802,
      "learning_rate": 5e-06,
      "loss": 0.4828,
      "step": 569
    },
    {
      "epoch": 0.9878682842287695,
      "grad_norm": 7.394272015315274,
      "learning_rate": 5e-06,
      "loss": 0.4712,
      "step": 570
    },
    {
      "epoch": 0.9896013864818024,
      "grad_norm": 7.444201359488822,
      "learning_rate": 5e-06,
      "loss": 0.4331,
      "step": 571
    },
    {
      "epoch": 0.9913344887348353,
      "grad_norm": 11.567383760618894,
      "learning_rate": 5e-06,
      "loss": 0.4054,
      "step": 572
    },
    {
      "epoch": 0.9930675909878682,
      "grad_norm": 7.654569729523706,
      "learning_rate": 5e-06,
      "loss": 0.2859,
      "step": 573
    },
    {
      "epoch": 0.9948006932409013,
      "grad_norm": 9.84112920841301,
      "learning_rate": 5e-06,
      "loss": 0.4125,
      "step": 574
    },
    {
      "epoch": 0.9965337954939342,
      "grad_norm": 6.358432859322623,
      "learning_rate": 5e-06,
      "loss": 0.3863,
      "step": 575
    },
    {
      "epoch": 0.9982668977469671,
      "grad_norm": 7.052087450425389,
      "learning_rate": 5e-06,
      "loss": 0.4324,
      "step": 576
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.120651253988116,
      "learning_rate": 5e-06,
      "loss": 0.4318,
      "step": 577
    },
    {
      "epoch": 1.001733102253033,
      "grad_norm": 7.501441308448768,
      "learning_rate": 5e-06,
      "loss": 0.2819,
      "step": 578
    },
    {
      "epoch": 1.0034662045060658,
      "grad_norm": 9.02922019455597,
      "learning_rate": 5e-06,
      "loss": 0.4184,
      "step": 579
    },
    {
      "epoch": 1.0051993067590987,
      "grad_norm": 10.057929286305423,
      "learning_rate": 5e-06,
      "loss": 0.3214,
      "step": 580
    },
    {
      "epoch": 1.0069324090121317,
      "grad_norm": 9.03077099859705,
      "learning_rate": 5e-06,
      "loss": 0.3122,
      "step": 581
    },
    {
      "epoch": 1.0086655112651646,
      "grad_norm": 4.7266605587710435,
      "learning_rate": 5e-06,
      "loss": 0.3056,
      "step": 582
    },
    {
      "epoch": 1.0103986135181975,
      "grad_norm": 8.786743183564989,
      "learning_rate": 5e-06,
      "loss": 0.3692,
      "step": 583
    },
    {
      "epoch": 1.0121317157712304,
      "grad_norm": 5.609674727313933,
      "learning_rate": 5e-06,
      "loss": 0.2966,
      "step": 584
    },
    {
      "epoch": 1.0138648180242635,
      "grad_norm": 7.126023871489697,
      "learning_rate": 5e-06,
      "loss": 0.2932,
      "step": 585
    },
    {
      "epoch": 1.0155979202772965,
      "grad_norm": 8.030791154511666,
      "learning_rate": 5e-06,
      "loss": 0.2728,
      "step": 586
    },
    {
      "epoch": 1.0173310225303294,
      "grad_norm": 3.9149905872840804,
      "learning_rate": 5e-06,
      "loss": 0.2029,
      "step": 587
    },
    {
      "epoch": 1.0190641247833623,
      "grad_norm": 8.884640897192767,
      "learning_rate": 5e-06,
      "loss": 0.2646,
      "step": 588
    },
    {
      "epoch": 1.0207972270363952,
      "grad_norm": 18.088113226312206,
      "learning_rate": 5e-06,
      "loss": 0.4974,
      "step": 589
    },
    {
      "epoch": 1.022530329289428,
      "grad_norm": 7.981489462609598,
      "learning_rate": 5e-06,
      "loss": 0.1916,
      "step": 590
    },
    {
      "epoch": 1.024263431542461,
      "grad_norm": 5.631280402335384,
      "learning_rate": 5e-06,
      "loss": 0.2265,
      "step": 591
    },
    {
      "epoch": 1.025996533795494,
      "grad_norm": 12.228594080416348,
      "learning_rate": 5e-06,
      "loss": 0.2805,
      "step": 592
    },
    {
      "epoch": 1.0277296360485269,
      "grad_norm": 10.497896619564255,
      "learning_rate": 5e-06,
      "loss": 0.3394,
      "step": 593
    },
    {
      "epoch": 1.0294627383015598,
      "grad_norm": 8.379257714073914,
      "learning_rate": 5e-06,
      "loss": 0.2234,
      "step": 594
    },
    {
      "epoch": 1.0311958405545927,
      "grad_norm": 9.17172942947725,
      "learning_rate": 5e-06,
      "loss": 0.3336,
      "step": 595
    },
    {
      "epoch": 1.0329289428076256,
      "grad_norm": 9.670647908451548,
      "learning_rate": 5e-06,
      "loss": 0.2298,
      "step": 596
    },
    {
      "epoch": 1.0346620450606585,
      "grad_norm": 11.990533909813308,
      "learning_rate": 5e-06,
      "loss": 0.271,
      "step": 597
    },
    {
      "epoch": 1.0363951473136914,
      "grad_norm": 8.742174409340064,
      "learning_rate": 5e-06,
      "loss": 0.3772,
      "step": 598
    },
    {
      "epoch": 1.0381282495667243,
      "grad_norm": 12.558206764060952,
      "learning_rate": 5e-06,
      "loss": 0.353,
      "step": 599
    },
    {
      "epoch": 1.0398613518197575,
      "grad_norm": 7.8011937230553805,
      "learning_rate": 5e-06,
      "loss": 0.2546,
      "step": 600
    },
    {
      "epoch": 1.0415944540727904,
      "grad_norm": 15.785076984148915,
      "learning_rate": 5e-06,
      "loss": 0.3194,
      "step": 601
    },
    {
      "epoch": 1.0433275563258233,
      "grad_norm": 5.825068513524614,
      "learning_rate": 5e-06,
      "loss": 0.2954,
      "step": 602
    },
    {
      "epoch": 1.0450606585788562,
      "grad_norm": 5.991055815661957,
      "learning_rate": 5e-06,
      "loss": 0.3049,
      "step": 603
    },
    {
      "epoch": 1.0467937608318891,
      "grad_norm": 9.663049174465819,
      "learning_rate": 5e-06,
      "loss": 0.259,
      "step": 604
    },
    {
      "epoch": 1.048526863084922,
      "grad_norm": 5.822905056509091,
      "learning_rate": 5e-06,
      "loss": 0.2388,
      "step": 605
    },
    {
      "epoch": 1.050259965337955,
      "grad_norm": 8.425367434495067,
      "learning_rate": 5e-06,
      "loss": 0.3043,
      "step": 606
    },
    {
      "epoch": 1.0519930675909879,
      "grad_norm": 12.099041710099787,
      "learning_rate": 5e-06,
      "loss": 0.2891,
      "step": 607
    },
    {
      "epoch": 1.0537261698440208,
      "grad_norm": 10.823027144239333,
      "learning_rate": 5e-06,
      "loss": 0.3482,
      "step": 608
    },
    {
      "epoch": 1.0554592720970537,
      "grad_norm": 12.240567059315257,
      "learning_rate": 5e-06,
      "loss": 0.4219,
      "step": 609
    },
    {
      "epoch": 1.0571923743500866,
      "grad_norm": 11.192515351892252,
      "learning_rate": 5e-06,
      "loss": 0.3684,
      "step": 610
    },
    {
      "epoch": 1.0589254766031195,
      "grad_norm": 7.942400283259556,
      "learning_rate": 5e-06,
      "loss": 0.3886,
      "step": 611
    },
    {
      "epoch": 1.0606585788561524,
      "grad_norm": 9.385612482927659,
      "learning_rate": 5e-06,
      "loss": 0.2483,
      "step": 612
    },
    {
      "epoch": 1.0623916811091854,
      "grad_norm": 7.099582778062285,
      "learning_rate": 5e-06,
      "loss": 0.2223,
      "step": 613
    },
    {
      "epoch": 1.0641247833622183,
      "grad_norm": 10.184948560746815,
      "learning_rate": 5e-06,
      "loss": 0.3243,
      "step": 614
    },
    {
      "epoch": 1.0658578856152512,
      "grad_norm": 9.916114300215655,
      "learning_rate": 5e-06,
      "loss": 0.2715,
      "step": 615
    },
    {
      "epoch": 1.0675909878682843,
      "grad_norm": 10.29081304513316,
      "learning_rate": 5e-06,
      "loss": 0.2443,
      "step": 616
    },
    {
      "epoch": 1.0693240901213172,
      "grad_norm": 5.898802891207368,
      "learning_rate": 5e-06,
      "loss": 0.3013,
      "step": 617
    },
    {
      "epoch": 1.0710571923743502,
      "grad_norm": 5.445845415043095,
      "learning_rate": 5e-06,
      "loss": 0.2876,
      "step": 618
    },
    {
      "epoch": 1.072790294627383,
      "grad_norm": 9.302893682697983,
      "learning_rate": 5e-06,
      "loss": 0.239,
      "step": 619
    },
    {
      "epoch": 1.074523396880416,
      "grad_norm": 9.502959744021037,
      "learning_rate": 5e-06,
      "loss": 0.2433,
      "step": 620
    },
    {
      "epoch": 1.076256499133449,
      "grad_norm": 6.369166379126097,
      "learning_rate": 5e-06,
      "loss": 0.2406,
      "step": 621
    },
    {
      "epoch": 1.0779896013864818,
      "grad_norm": 13.252632311491697,
      "learning_rate": 5e-06,
      "loss": 0.2765,
      "step": 622
    },
    {
      "epoch": 1.0797227036395147,
      "grad_norm": 12.186756913444315,
      "learning_rate": 5e-06,
      "loss": 0.3019,
      "step": 623
    },
    {
      "epoch": 1.0814558058925476,
      "grad_norm": 5.243958948494087,
      "learning_rate": 5e-06,
      "loss": 0.1574,
      "step": 624
    },
    {
      "epoch": 1.0831889081455806,
      "grad_norm": 9.551546093480306,
      "learning_rate": 5e-06,
      "loss": 0.1671,
      "step": 625
    },
    {
      "epoch": 1.0849220103986135,
      "grad_norm": 10.88653233604305,
      "learning_rate": 5e-06,
      "loss": 0.4014,
      "step": 626
    },
    {
      "epoch": 1.0866551126516464,
      "grad_norm": 6.115405606333931,
      "learning_rate": 5e-06,
      "loss": 0.2532,
      "step": 627
    },
    {
      "epoch": 1.0883882149046793,
      "grad_norm": 8.674451510544575,
      "learning_rate": 5e-06,
      "loss": 0.2072,
      "step": 628
    },
    {
      "epoch": 1.0901213171577122,
      "grad_norm": 11.890079601075927,
      "learning_rate": 5e-06,
      "loss": 0.2569,
      "step": 629
    },
    {
      "epoch": 1.0918544194107453,
      "grad_norm": 10.76495917355476,
      "learning_rate": 5e-06,
      "loss": 0.2695,
      "step": 630
    },
    {
      "epoch": 1.0935875216637783,
      "grad_norm": 10.744260453396953,
      "learning_rate": 5e-06,
      "loss": 0.439,
      "step": 631
    },
    {
      "epoch": 1.0953206239168112,
      "grad_norm": 17.03058520778224,
      "learning_rate": 5e-06,
      "loss": 0.5291,
      "step": 632
    },
    {
      "epoch": 1.097053726169844,
      "grad_norm": 7.505446173247896,
      "learning_rate": 5e-06,
      "loss": 0.2261,
      "step": 633
    },
    {
      "epoch": 1.098786828422877,
      "grad_norm": 11.638410155273908,
      "learning_rate": 5e-06,
      "loss": 0.3367,
      "step": 634
    },
    {
      "epoch": 1.10051993067591,
      "grad_norm": 12.256947940949592,
      "learning_rate": 5e-06,
      "loss": 0.4265,
      "step": 635
    },
    {
      "epoch": 1.1022530329289428,
      "grad_norm": 6.0329884607934074,
      "learning_rate": 5e-06,
      "loss": 0.2235,
      "step": 636
    },
    {
      "epoch": 1.1039861351819757,
      "grad_norm": 8.991793387730654,
      "learning_rate": 5e-06,
      "loss": 0.2762,
      "step": 637
    },
    {
      "epoch": 1.1057192374350087,
      "grad_norm": 11.77826424159867,
      "learning_rate": 5e-06,
      "loss": 0.2825,
      "step": 638
    },
    {
      "epoch": 1.1074523396880416,
      "grad_norm": 14.903094775472601,
      "learning_rate": 5e-06,
      "loss": 0.3297,
      "step": 639
    },
    {
      "epoch": 1.1091854419410745,
      "grad_norm": 9.097929633075783,
      "learning_rate": 5e-06,
      "loss": 0.2998,
      "step": 640
    },
    {
      "epoch": 1.1109185441941074,
      "grad_norm": 15.595916780980136,
      "learning_rate": 5e-06,
      "loss": 0.3749,
      "step": 641
    },
    {
      "epoch": 1.1126516464471403,
      "grad_norm": 7.502036771777226,
      "learning_rate": 5e-06,
      "loss": 0.2824,
      "step": 642
    },
    {
      "epoch": 1.1143847487001732,
      "grad_norm": 7.964959653092452,
      "learning_rate": 5e-06,
      "loss": 0.3171,
      "step": 643
    },
    {
      "epoch": 1.1161178509532061,
      "grad_norm": 6.961928514349743,
      "learning_rate": 5e-06,
      "loss": 0.3448,
      "step": 644
    },
    {
      "epoch": 1.117850953206239,
      "grad_norm": 6.838014193488601,
      "learning_rate": 5e-06,
      "loss": 0.221,
      "step": 645
    },
    {
      "epoch": 1.1195840554592722,
      "grad_norm": 10.654036037423506,
      "learning_rate": 5e-06,
      "loss": 0.2493,
      "step": 646
    },
    {
      "epoch": 1.1213171577123051,
      "grad_norm": 6.731303849000456,
      "learning_rate": 5e-06,
      "loss": 0.3064,
      "step": 647
    },
    {
      "epoch": 1.123050259965338,
      "grad_norm": 11.902796927908636,
      "learning_rate": 5e-06,
      "loss": 0.3368,
      "step": 648
    },
    {
      "epoch": 1.124783362218371,
      "grad_norm": 7.461497596368656,
      "learning_rate": 5e-06,
      "loss": 0.303,
      "step": 649
    },
    {
      "epoch": 1.1265164644714039,
      "grad_norm": 9.372074331902988,
      "learning_rate": 5e-06,
      "loss": 0.2386,
      "step": 650
    },
    {
      "epoch": 1.1282495667244368,
      "grad_norm": 7.5010141322820685,
      "learning_rate": 5e-06,
      "loss": 0.2764,
      "step": 651
    },
    {
      "epoch": 1.1299826689774697,
      "grad_norm": 6.905341593519712,
      "learning_rate": 5e-06,
      "loss": 0.2339,
      "step": 652
    },
    {
      "epoch": 1.1317157712305026,
      "grad_norm": 6.355526971345184,
      "learning_rate": 5e-06,
      "loss": 0.3197,
      "step": 653
    },
    {
      "epoch": 1.1334488734835355,
      "grad_norm": 11.187747547671501,
      "learning_rate": 5e-06,
      "loss": 0.2109,
      "step": 654
    },
    {
      "epoch": 1.1351819757365684,
      "grad_norm": 7.757631051975482,
      "learning_rate": 5e-06,
      "loss": 0.3027,
      "step": 655
    },
    {
      "epoch": 1.1369150779896013,
      "grad_norm": 9.109449121294714,
      "learning_rate": 5e-06,
      "loss": 0.3109,
      "step": 656
    },
    {
      "epoch": 1.1386481802426343,
      "grad_norm": 12.397492284385832,
      "learning_rate": 5e-06,
      "loss": 0.472,
      "step": 657
    },
    {
      "epoch": 1.1403812824956672,
      "grad_norm": 9.071606079699139,
      "learning_rate": 5e-06,
      "loss": 0.262,
      "step": 658
    },
    {
      "epoch": 1.1421143847487,
      "grad_norm": 6.537382468339566,
      "learning_rate": 5e-06,
      "loss": 0.2275,
      "step": 659
    },
    {
      "epoch": 1.1438474870017332,
      "grad_norm": 7.590735068530269,
      "learning_rate": 5e-06,
      "loss": 0.2414,
      "step": 660
    },
    {
      "epoch": 1.1455805892547661,
      "grad_norm": 6.460782806869386,
      "learning_rate": 5e-06,
      "loss": 0.3054,
      "step": 661
    },
    {
      "epoch": 1.147313691507799,
      "grad_norm": 7.220172164005569,
      "learning_rate": 5e-06,
      "loss": 0.3101,
      "step": 662
    },
    {
      "epoch": 1.149046793760832,
      "grad_norm": 8.386156536814761,
      "learning_rate": 5e-06,
      "loss": 0.1688,
      "step": 663
    },
    {
      "epoch": 1.1507798960138649,
      "grad_norm": 6.607716309394605,
      "learning_rate": 5e-06,
      "loss": 0.2432,
      "step": 664
    },
    {
      "epoch": 1.1525129982668978,
      "grad_norm": 16.060067909258667,
      "learning_rate": 5e-06,
      "loss": 0.4351,
      "step": 665
    },
    {
      "epoch": 1.1542461005199307,
      "grad_norm": 17.423571421050653,
      "learning_rate": 5e-06,
      "loss": 0.3173,
      "step": 666
    },
    {
      "epoch": 1.1559792027729636,
      "grad_norm": 13.211639728527501,
      "learning_rate": 5e-06,
      "loss": 0.3469,
      "step": 667
    },
    {
      "epoch": 1.1577123050259965,
      "grad_norm": 12.676995800826775,
      "learning_rate": 5e-06,
      "loss": 0.2804,
      "step": 668
    },
    {
      "epoch": 1.1594454072790294,
      "grad_norm": 9.291194841068991,
      "learning_rate": 5e-06,
      "loss": 0.2866,
      "step": 669
    },
    {
      "epoch": 1.1611785095320624,
      "grad_norm": 6.478980778258941,
      "learning_rate": 5e-06,
      "loss": 0.2166,
      "step": 670
    },
    {
      "epoch": 1.1629116117850953,
      "grad_norm": 8.511269671781129,
      "learning_rate": 5e-06,
      "loss": 0.3013,
      "step": 671
    },
    {
      "epoch": 1.1646447140381282,
      "grad_norm": 6.2720077614638905,
      "learning_rate": 5e-06,
      "loss": 0.2357,
      "step": 672
    },
    {
      "epoch": 1.166377816291161,
      "grad_norm": 7.686534045559476,
      "learning_rate": 5e-06,
      "loss": 0.248,
      "step": 673
    },
    {
      "epoch": 1.168110918544194,
      "grad_norm": 6.090034486770073,
      "learning_rate": 5e-06,
      "loss": 0.2934,
      "step": 674
    },
    {
      "epoch": 1.169844020797227,
      "grad_norm": 8.774198368331772,
      "learning_rate": 5e-06,
      "loss": 0.3813,
      "step": 675
    },
    {
      "epoch": 1.1715771230502598,
      "grad_norm": 5.696978708600726,
      "learning_rate": 5e-06,
      "loss": 0.2271,
      "step": 676
    },
    {
      "epoch": 1.173310225303293,
      "grad_norm": 13.2294722554313,
      "learning_rate": 5e-06,
      "loss": 0.3268,
      "step": 677
    },
    {
      "epoch": 1.175043327556326,
      "grad_norm": 8.082679276370428,
      "learning_rate": 5e-06,
      "loss": 0.3299,
      "step": 678
    },
    {
      "epoch": 1.1767764298093588,
      "grad_norm": 6.8335746474417185,
      "learning_rate": 5e-06,
      "loss": 0.2482,
      "step": 679
    },
    {
      "epoch": 1.1785095320623917,
      "grad_norm": 8.597079876892384,
      "learning_rate": 5e-06,
      "loss": 0.2554,
      "step": 680
    },
    {
      "epoch": 1.1802426343154246,
      "grad_norm": 11.508658425290065,
      "learning_rate": 5e-06,
      "loss": 0.3749,
      "step": 681
    },
    {
      "epoch": 1.1819757365684576,
      "grad_norm": 9.664251968635817,
      "learning_rate": 5e-06,
      "loss": 0.3386,
      "step": 682
    },
    {
      "epoch": 1.1837088388214905,
      "grad_norm": 11.478895642265227,
      "learning_rate": 5e-06,
      "loss": 0.1992,
      "step": 683
    },
    {
      "epoch": 1.1854419410745234,
      "grad_norm": 7.120345619899793,
      "learning_rate": 5e-06,
      "loss": 0.34,
      "step": 684
    },
    {
      "epoch": 1.1871750433275563,
      "grad_norm": 18.096002362778137,
      "learning_rate": 5e-06,
      "loss": 0.3199,
      "step": 685
    },
    {
      "epoch": 1.1889081455805892,
      "grad_norm": 6.799474740778752,
      "learning_rate": 5e-06,
      "loss": 0.1704,
      "step": 686
    },
    {
      "epoch": 1.1906412478336221,
      "grad_norm": 5.168245176965632,
      "learning_rate": 5e-06,
      "loss": 0.2055,
      "step": 687
    },
    {
      "epoch": 1.192374350086655,
      "grad_norm": 6.905906979426822,
      "learning_rate": 5e-06,
      "loss": 0.317,
      "step": 688
    },
    {
      "epoch": 1.194107452339688,
      "grad_norm": 10.825616719427906,
      "learning_rate": 5e-06,
      "loss": 0.3504,
      "step": 689
    },
    {
      "epoch": 1.195840554592721,
      "grad_norm": 5.840241873448467,
      "learning_rate": 5e-06,
      "loss": 0.1777,
      "step": 690
    },
    {
      "epoch": 1.197573656845754,
      "grad_norm": 6.6768725793897135,
      "learning_rate": 5e-06,
      "loss": 0.2329,
      "step": 691
    },
    {
      "epoch": 1.199306759098787,
      "grad_norm": 8.220526093588603,
      "learning_rate": 5e-06,
      "loss": 0.3015,
      "step": 692
    },
    {
      "epoch": 1.2010398613518198,
      "grad_norm": 7.2261174905212,
      "learning_rate": 5e-06,
      "loss": 0.1737,
      "step": 693
    },
    {
      "epoch": 1.2027729636048528,
      "grad_norm": 11.182228252845247,
      "learning_rate": 5e-06,
      "loss": 0.2617,
      "step": 694
    },
    {
      "epoch": 1.2045060658578857,
      "grad_norm": 13.485278544648525,
      "learning_rate": 5e-06,
      "loss": 0.3997,
      "step": 695
    },
    {
      "epoch": 1.2062391681109186,
      "grad_norm": 8.528454892105586,
      "learning_rate": 5e-06,
      "loss": 0.1684,
      "step": 696
    },
    {
      "epoch": 1.2079722703639515,
      "grad_norm": 10.3454555246353,
      "learning_rate": 5e-06,
      "loss": 0.3805,
      "step": 697
    },
    {
      "epoch": 1.2097053726169844,
      "grad_norm": 10.29347202873015,
      "learning_rate": 5e-06,
      "loss": 0.353,
      "step": 698
    },
    {
      "epoch": 1.2114384748700173,
      "grad_norm": 5.802633463759889,
      "learning_rate": 5e-06,
      "loss": 0.1566,
      "step": 699
    },
    {
      "epoch": 1.2131715771230502,
      "grad_norm": 13.37830860891434,
      "learning_rate": 5e-06,
      "loss": 0.4165,
      "step": 700
    },
    {
      "epoch": 1.2149046793760832,
      "grad_norm": 10.842974632311611,
      "learning_rate": 5e-06,
      "loss": 0.2341,
      "step": 701
    },
    {
      "epoch": 1.216637781629116,
      "grad_norm": 7.9847373805515325,
      "learning_rate": 5e-06,
      "loss": 0.3476,
      "step": 702
    },
    {
      "epoch": 1.218370883882149,
      "grad_norm": 9.43563561623815,
      "learning_rate": 5e-06,
      "loss": 0.2583,
      "step": 703
    },
    {
      "epoch": 1.220103986135182,
      "grad_norm": 10.147525173700181,
      "learning_rate": 5e-06,
      "loss": 0.2844,
      "step": 704
    },
    {
      "epoch": 1.2218370883882148,
      "grad_norm": 6.8251589131281865,
      "learning_rate": 5e-06,
      "loss": 0.2213,
      "step": 705
    },
    {
      "epoch": 1.2235701906412477,
      "grad_norm": 8.584883565318643,
      "learning_rate": 5e-06,
      "loss": 0.2666,
      "step": 706
    },
    {
      "epoch": 1.2253032928942806,
      "grad_norm": 12.935223512727514,
      "learning_rate": 5e-06,
      "loss": 0.2992,
      "step": 707
    },
    {
      "epoch": 1.2270363951473138,
      "grad_norm": 11.302677168239107,
      "learning_rate": 5e-06,
      "loss": 0.3784,
      "step": 708
    },
    {
      "epoch": 1.2287694974003467,
      "grad_norm": 7.394262729128108,
      "learning_rate": 5e-06,
      "loss": 0.2722,
      "step": 709
    },
    {
      "epoch": 1.2305025996533796,
      "grad_norm": 11.27326898587671,
      "learning_rate": 5e-06,
      "loss": 0.3206,
      "step": 710
    },
    {
      "epoch": 1.2322357019064125,
      "grad_norm": 8.487013376749733,
      "learning_rate": 5e-06,
      "loss": 0.3238,
      "step": 711
    },
    {
      "epoch": 1.2339688041594454,
      "grad_norm": 7.568895898321558,
      "learning_rate": 5e-06,
      "loss": 0.3202,
      "step": 712
    },
    {
      "epoch": 1.2357019064124783,
      "grad_norm": 7.616508383187921,
      "learning_rate": 5e-06,
      "loss": 0.2303,
      "step": 713
    },
    {
      "epoch": 1.2374350086655113,
      "grad_norm": 5.784641570095975,
      "learning_rate": 5e-06,
      "loss": 0.2255,
      "step": 714
    },
    {
      "epoch": 1.2391681109185442,
      "grad_norm": 7.400002979587264,
      "learning_rate": 5e-06,
      "loss": 0.2053,
      "step": 715
    },
    {
      "epoch": 1.240901213171577,
      "grad_norm": 7.953016119251011,
      "learning_rate": 5e-06,
      "loss": 0.3047,
      "step": 716
    },
    {
      "epoch": 1.24263431542461,
      "grad_norm": 10.030403458000752,
      "learning_rate": 5e-06,
      "loss": 0.2843,
      "step": 717
    },
    {
      "epoch": 1.244367417677643,
      "grad_norm": 7.475209782441153,
      "learning_rate": 5e-06,
      "loss": 0.2156,
      "step": 718
    },
    {
      "epoch": 1.2461005199306758,
      "grad_norm": 5.6241130553449725,
      "learning_rate": 5e-06,
      "loss": 0.2119,
      "step": 719
    },
    {
      "epoch": 1.2478336221837087,
      "grad_norm": 7.875212409167257,
      "learning_rate": 5e-06,
      "loss": 0.2308,
      "step": 720
    },
    {
      "epoch": 1.2495667244367419,
      "grad_norm": 6.337375388340519,
      "learning_rate": 5e-06,
      "loss": 0.2231,
      "step": 721
    },
    {
      "epoch": 1.2512998266897748,
      "grad_norm": 6.612749974045464,
      "learning_rate": 5e-06,
      "loss": 0.2407,
      "step": 722
    },
    {
      "epoch": 1.2530329289428077,
      "grad_norm": 12.802031618059873,
      "learning_rate": 5e-06,
      "loss": 0.3598,
      "step": 723
    },
    {
      "epoch": 1.2547660311958406,
      "grad_norm": 7.469617466049471,
      "learning_rate": 5e-06,
      "loss": 0.273,
      "step": 724
    },
    {
      "epoch": 1.2564991334488735,
      "grad_norm": 7.558821358814657,
      "learning_rate": 5e-06,
      "loss": 0.2474,
      "step": 725
    },
    {
      "epoch": 1.2582322357019065,
      "grad_norm": 9.208507783598053,
      "learning_rate": 5e-06,
      "loss": 0.3295,
      "step": 726
    },
    {
      "epoch": 1.2599653379549394,
      "grad_norm": 7.228938059714677,
      "learning_rate": 5e-06,
      "loss": 0.2688,
      "step": 727
    },
    {
      "epoch": 1.2616984402079723,
      "grad_norm": 13.843825501522387,
      "learning_rate": 5e-06,
      "loss": 0.3899,
      "step": 728
    },
    {
      "epoch": 1.2634315424610052,
      "grad_norm": 11.201564366216367,
      "learning_rate": 5e-06,
      "loss": 0.3793,
      "step": 729
    },
    {
      "epoch": 1.2651646447140381,
      "grad_norm": 9.939184657837767,
      "learning_rate": 5e-06,
      "loss": 0.4001,
      "step": 730
    },
    {
      "epoch": 1.266897746967071,
      "grad_norm": 12.933698749951315,
      "learning_rate": 5e-06,
      "loss": 0.4011,
      "step": 731
    },
    {
      "epoch": 1.268630849220104,
      "grad_norm": 7.147753073023712,
      "learning_rate": 5e-06,
      "loss": 0.3006,
      "step": 732
    },
    {
      "epoch": 1.2703639514731369,
      "grad_norm": 7.5635266828754135,
      "learning_rate": 5e-06,
      "loss": 0.3442,
      "step": 733
    },
    {
      "epoch": 1.2720970537261698,
      "grad_norm": 6.8443597669991885,
      "learning_rate": 5e-06,
      "loss": 0.1859,
      "step": 734
    },
    {
      "epoch": 1.2738301559792027,
      "grad_norm": 7.763432091422889,
      "learning_rate": 5e-06,
      "loss": 0.2194,
      "step": 735
    },
    {
      "epoch": 1.2755632582322356,
      "grad_norm": 9.798633667033513,
      "learning_rate": 5e-06,
      "loss": 0.262,
      "step": 736
    },
    {
      "epoch": 1.2772963604852685,
      "grad_norm": 7.626283381285403,
      "learning_rate": 5e-06,
      "loss": 0.3678,
      "step": 737
    },
    {
      "epoch": 1.2790294627383014,
      "grad_norm": 6.194470111392701,
      "learning_rate": 5e-06,
      "loss": 0.1872,
      "step": 738
    },
    {
      "epoch": 1.2807625649913346,
      "grad_norm": 8.737554363905671,
      "learning_rate": 5e-06,
      "loss": 0.2551,
      "step": 739
    },
    {
      "epoch": 1.2824956672443675,
      "grad_norm": 12.82687420154367,
      "learning_rate": 5e-06,
      "loss": 0.2764,
      "step": 740
    },
    {
      "epoch": 1.2842287694974004,
      "grad_norm": 15.030711967371797,
      "learning_rate": 5e-06,
      "loss": 0.4273,
      "step": 741
    },
    {
      "epoch": 1.2859618717504333,
      "grad_norm": 7.085378942377047,
      "learning_rate": 5e-06,
      "loss": 0.2166,
      "step": 742
    },
    {
      "epoch": 1.2876949740034662,
      "grad_norm": 7.158694720073964,
      "learning_rate": 5e-06,
      "loss": 0.3896,
      "step": 743
    },
    {
      "epoch": 1.2894280762564991,
      "grad_norm": 8.58748617601462,
      "learning_rate": 5e-06,
      "loss": 0.2985,
      "step": 744
    },
    {
      "epoch": 1.291161178509532,
      "grad_norm": 6.475595289283135,
      "learning_rate": 5e-06,
      "loss": 0.3405,
      "step": 745
    },
    {
      "epoch": 1.292894280762565,
      "grad_norm": 6.402786137190865,
      "learning_rate": 5e-06,
      "loss": 0.3157,
      "step": 746
    },
    {
      "epoch": 1.2946273830155979,
      "grad_norm": 7.24223701270831,
      "learning_rate": 5e-06,
      "loss": 0.2698,
      "step": 747
    },
    {
      "epoch": 1.2963604852686308,
      "grad_norm": 6.210162343530788,
      "learning_rate": 5e-06,
      "loss": 0.2731,
      "step": 748
    },
    {
      "epoch": 1.2980935875216637,
      "grad_norm": 3.8714051337932767,
      "learning_rate": 5e-06,
      "loss": 0.2028,
      "step": 749
    },
    {
      "epoch": 1.2998266897746968,
      "grad_norm": 7.048587836199471,
      "learning_rate": 5e-06,
      "loss": 0.2913,
      "step": 750
    },
    {
      "epoch": 1.3015597920277298,
      "grad_norm": 6.223157997760293,
      "learning_rate": 5e-06,
      "loss": 0.2776,
      "step": 751
    },
    {
      "epoch": 1.3032928942807627,
      "grad_norm": 6.829082500442615,
      "learning_rate": 5e-06,
      "loss": 0.2796,
      "step": 752
    },
    {
      "epoch": 1.3050259965337956,
      "grad_norm": 6.240094995509164,
      "learning_rate": 5e-06,
      "loss": 0.2146,
      "step": 753
    },
    {
      "epoch": 1.3067590987868285,
      "grad_norm": 7.1197889579613385,
      "learning_rate": 5e-06,
      "loss": 0.2837,
      "step": 754
    },
    {
      "epoch": 1.3084922010398614,
      "grad_norm": 11.924374379635315,
      "learning_rate": 5e-06,
      "loss": 0.3137,
      "step": 755
    },
    {
      "epoch": 1.3102253032928943,
      "grad_norm": 7.061463136432652,
      "learning_rate": 5e-06,
      "loss": 0.2901,
      "step": 756
    },
    {
      "epoch": 1.3119584055459272,
      "grad_norm": 12.5239254671387,
      "learning_rate": 5e-06,
      "loss": 0.4602,
      "step": 757
    },
    {
      "epoch": 1.3136915077989602,
      "grad_norm": 10.015430371183234,
      "learning_rate": 5e-06,
      "loss": 0.2467,
      "step": 758
    },
    {
      "epoch": 1.315424610051993,
      "grad_norm": 8.276225167015257,
      "learning_rate": 5e-06,
      "loss": 0.2155,
      "step": 759
    },
    {
      "epoch": 1.317157712305026,
      "grad_norm": 7.535916857104865,
      "learning_rate": 5e-06,
      "loss": 0.2098,
      "step": 760
    },
    {
      "epoch": 1.318890814558059,
      "grad_norm": 9.533638676730781,
      "learning_rate": 5e-06,
      "loss": 0.2198,
      "step": 761
    },
    {
      "epoch": 1.3206239168110918,
      "grad_norm": 7.448475152975418,
      "learning_rate": 5e-06,
      "loss": 0.1991,
      "step": 762
    },
    {
      "epoch": 1.3223570190641247,
      "grad_norm": 9.742288914743796,
      "learning_rate": 5e-06,
      "loss": 0.243,
      "step": 763
    },
    {
      "epoch": 1.3240901213171576,
      "grad_norm": 14.730872214210747,
      "learning_rate": 5e-06,
      "loss": 0.4395,
      "step": 764
    },
    {
      "epoch": 1.3258232235701906,
      "grad_norm": 11.64251243087698,
      "learning_rate": 5e-06,
      "loss": 0.2386,
      "step": 765
    },
    {
      "epoch": 1.3275563258232235,
      "grad_norm": 10.684864812414348,
      "learning_rate": 5e-06,
      "loss": 0.2488,
      "step": 766
    },
    {
      "epoch": 1.3292894280762564,
      "grad_norm": 5.988719508587713,
      "learning_rate": 5e-06,
      "loss": 0.2205,
      "step": 767
    },
    {
      "epoch": 1.3310225303292893,
      "grad_norm": 10.094100331937982,
      "learning_rate": 5e-06,
      "loss": 0.3281,
      "step": 768
    },
    {
      "epoch": 1.3327556325823224,
      "grad_norm": 10.802499015753325,
      "learning_rate": 5e-06,
      "loss": 0.2856,
      "step": 769
    },
    {
      "epoch": 1.3344887348353554,
      "grad_norm": 11.09357634932314,
      "learning_rate": 5e-06,
      "loss": 0.2688,
      "step": 770
    },
    {
      "epoch": 1.3362218370883883,
      "grad_norm": 9.91396822933411,
      "learning_rate": 5e-06,
      "loss": 0.3399,
      "step": 771
    },
    {
      "epoch": 1.3379549393414212,
      "grad_norm": 6.2216529396546765,
      "learning_rate": 5e-06,
      "loss": 0.2659,
      "step": 772
    },
    {
      "epoch": 1.339688041594454,
      "grad_norm": 5.78897848338578,
      "learning_rate": 5e-06,
      "loss": 0.268,
      "step": 773
    },
    {
      "epoch": 1.341421143847487,
      "grad_norm": 10.187143518494587,
      "learning_rate": 5e-06,
      "loss": 0.2587,
      "step": 774
    },
    {
      "epoch": 1.34315424610052,
      "grad_norm": 5.495590523115228,
      "learning_rate": 5e-06,
      "loss": 0.1837,
      "step": 775
    },
    {
      "epoch": 1.3448873483535528,
      "grad_norm": 10.003968786411331,
      "learning_rate": 5e-06,
      "loss": 0.2578,
      "step": 776
    },
    {
      "epoch": 1.3466204506065858,
      "grad_norm": 5.900491480642503,
      "learning_rate": 5e-06,
      "loss": 0.2292,
      "step": 777
    },
    {
      "epoch": 1.3483535528596187,
      "grad_norm": 11.715941069606162,
      "learning_rate": 5e-06,
      "loss": 0.411,
      "step": 778
    },
    {
      "epoch": 1.3500866551126516,
      "grad_norm": 6.287149723678387,
      "learning_rate": 5e-06,
      "loss": 0.2824,
      "step": 779
    },
    {
      "epoch": 1.3518197573656845,
      "grad_norm": 9.83319475324148,
      "learning_rate": 5e-06,
      "loss": 0.282,
      "step": 780
    },
    {
      "epoch": 1.3535528596187176,
      "grad_norm": 6.591437779570791,
      "learning_rate": 5e-06,
      "loss": 0.3049,
      "step": 781
    },
    {
      "epoch": 1.3552859618717505,
      "grad_norm": 9.194749113210365,
      "learning_rate": 5e-06,
      "loss": 0.2807,
      "step": 782
    },
    {
      "epoch": 1.3570190641247835,
      "grad_norm": 7.809465475119285,
      "learning_rate": 5e-06,
      "loss": 0.2979,
      "step": 783
    },
    {
      "epoch": 1.3587521663778164,
      "grad_norm": 9.237873273232573,
      "learning_rate": 5e-06,
      "loss": 0.3134,
      "step": 784
    },
    {
      "epoch": 1.3604852686308493,
      "grad_norm": 11.023938919872608,
      "learning_rate": 5e-06,
      "loss": 0.3444,
      "step": 785
    },
    {
      "epoch": 1.3622183708838822,
      "grad_norm": 8.501198179280285,
      "learning_rate": 5e-06,
      "loss": 0.3132,
      "step": 786
    },
    {
      "epoch": 1.3639514731369151,
      "grad_norm": 4.975738792579693,
      "learning_rate": 5e-06,
      "loss": 0.1769,
      "step": 787
    },
    {
      "epoch": 1.365684575389948,
      "grad_norm": 12.291507077662786,
      "learning_rate": 5e-06,
      "loss": 0.282,
      "step": 788
    },
    {
      "epoch": 1.367417677642981,
      "grad_norm": 12.136103314577928,
      "learning_rate": 5e-06,
      "loss": 0.3999,
      "step": 789
    },
    {
      "epoch": 1.3691507798960139,
      "grad_norm": 9.746846789545181,
      "learning_rate": 5e-06,
      "loss": 0.3453,
      "step": 790
    },
    {
      "epoch": 1.3708838821490468,
      "grad_norm": 13.432699224179242,
      "learning_rate": 5e-06,
      "loss": 0.3767,
      "step": 791
    },
    {
      "epoch": 1.3726169844020797,
      "grad_norm": 14.44453397543784,
      "learning_rate": 5e-06,
      "loss": 0.4026,
      "step": 792
    },
    {
      "epoch": 1.3743500866551126,
      "grad_norm": 9.89394444692055,
      "learning_rate": 5e-06,
      "loss": 0.2069,
      "step": 793
    },
    {
      "epoch": 1.3760831889081455,
      "grad_norm": 4.760026488671209,
      "learning_rate": 5e-06,
      "loss": 0.1305,
      "step": 794
    },
    {
      "epoch": 1.3778162911611784,
      "grad_norm": 7.176864052252329,
      "learning_rate": 5e-06,
      "loss": 0.1937,
      "step": 795
    },
    {
      "epoch": 1.3795493934142113,
      "grad_norm": 12.092819454065822,
      "learning_rate": 5e-06,
      "loss": 0.3307,
      "step": 796
    },
    {
      "epoch": 1.3812824956672443,
      "grad_norm": 6.065113251612274,
      "learning_rate": 5e-06,
      "loss": 0.2785,
      "step": 797
    },
    {
      "epoch": 1.3830155979202772,
      "grad_norm": 8.658851425438163,
      "learning_rate": 5e-06,
      "loss": 0.2367,
      "step": 798
    },
    {
      "epoch": 1.38474870017331,
      "grad_norm": 7.221304833598694,
      "learning_rate": 5e-06,
      "loss": 0.1811,
      "step": 799
    },
    {
      "epoch": 1.3864818024263432,
      "grad_norm": 7.834292116400045,
      "learning_rate": 5e-06,
      "loss": 0.2486,
      "step": 800
    },
    {
      "epoch": 1.3882149046793761,
      "grad_norm": 8.149198677594875,
      "learning_rate": 5e-06,
      "loss": 0.2208,
      "step": 801
    },
    {
      "epoch": 1.389948006932409,
      "grad_norm": 9.364148178994396,
      "learning_rate": 5e-06,
      "loss": 0.2513,
      "step": 802
    },
    {
      "epoch": 1.391681109185442,
      "grad_norm": 11.119290376154328,
      "learning_rate": 5e-06,
      "loss": 0.386,
      "step": 803
    },
    {
      "epoch": 1.3934142114384749,
      "grad_norm": 6.068615846889564,
      "learning_rate": 5e-06,
      "loss": 0.1878,
      "step": 804
    },
    {
      "epoch": 1.3951473136915078,
      "grad_norm": 8.994457657618245,
      "learning_rate": 5e-06,
      "loss": 0.3084,
      "step": 805
    },
    {
      "epoch": 1.3968804159445407,
      "grad_norm": 4.785158242984279,
      "learning_rate": 5e-06,
      "loss": 0.1673,
      "step": 806
    },
    {
      "epoch": 1.3986135181975736,
      "grad_norm": 8.349225365803965,
      "learning_rate": 5e-06,
      "loss": 0.2882,
      "step": 807
    },
    {
      "epoch": 1.4003466204506065,
      "grad_norm": 8.966837502971018,
      "learning_rate": 5e-06,
      "loss": 0.3149,
      "step": 808
    },
    {
      "epoch": 1.4020797227036395,
      "grad_norm": 11.109082096064006,
      "learning_rate": 5e-06,
      "loss": 0.2099,
      "step": 809
    },
    {
      "epoch": 1.4038128249566724,
      "grad_norm": 11.687258468008949,
      "learning_rate": 5e-06,
      "loss": 0.2155,
      "step": 810
    },
    {
      "epoch": 1.4055459272097055,
      "grad_norm": 9.892225862126493,
      "learning_rate": 5e-06,
      "loss": 0.375,
      "step": 811
    },
    {
      "epoch": 1.4072790294627384,
      "grad_norm": 11.235214350599025,
      "learning_rate": 5e-06,
      "loss": 0.2815,
      "step": 812
    },
    {
      "epoch": 1.4090121317157713,
      "grad_norm": 12.731455162766094,
      "learning_rate": 5e-06,
      "loss": 0.2729,
      "step": 813
    },
    {
      "epoch": 1.4107452339688042,
      "grad_norm": 6.671204580174192,
      "learning_rate": 5e-06,
      "loss": 0.2447,
      "step": 814
    },
    {
      "epoch": 1.4124783362218372,
      "grad_norm": 5.972201163479053,
      "learning_rate": 5e-06,
      "loss": 0.1502,
      "step": 815
    },
    {
      "epoch": 1.41421143847487,
      "grad_norm": 11.539679364224188,
      "learning_rate": 5e-06,
      "loss": 0.4281,
      "step": 816
    },
    {
      "epoch": 1.415944540727903,
      "grad_norm": 9.308954370375147,
      "learning_rate": 5e-06,
      "loss": 0.2087,
      "step": 817
    },
    {
      "epoch": 1.417677642980936,
      "grad_norm": 19.056955200238736,
      "learning_rate": 5e-06,
      "loss": 0.3333,
      "step": 818
    },
    {
      "epoch": 1.4194107452339688,
      "grad_norm": 8.910241875410584,
      "learning_rate": 5e-06,
      "loss": 0.2398,
      "step": 819
    },
    {
      "epoch": 1.4211438474870017,
      "grad_norm": 6.329495354614916,
      "learning_rate": 5e-06,
      "loss": 0.2822,
      "step": 820
    },
    {
      "epoch": 1.4228769497400346,
      "grad_norm": 8.826334597899743,
      "learning_rate": 5e-06,
      "loss": 0.2484,
      "step": 821
    },
    {
      "epoch": 1.4246100519930676,
      "grad_norm": 8.480625286441056,
      "learning_rate": 5e-06,
      "loss": 0.2703,
      "step": 822
    },
    {
      "epoch": 1.4263431542461005,
      "grad_norm": 6.677880900611654,
      "learning_rate": 5e-06,
      "loss": 0.1814,
      "step": 823
    },
    {
      "epoch": 1.4280762564991334,
      "grad_norm": 5.707631125963586,
      "learning_rate": 5e-06,
      "loss": 0.2067,
      "step": 824
    },
    {
      "epoch": 1.4298093587521663,
      "grad_norm": 8.162664353963645,
      "learning_rate": 5e-06,
      "loss": 0.2498,
      "step": 825
    },
    {
      "epoch": 1.4315424610051992,
      "grad_norm": 8.171732622295774,
      "learning_rate": 5e-06,
      "loss": 0.4091,
      "step": 826
    },
    {
      "epoch": 1.4332755632582321,
      "grad_norm": 8.435473389425809,
      "learning_rate": 5e-06,
      "loss": 0.1574,
      "step": 827
    },
    {
      "epoch": 1.435008665511265,
      "grad_norm": 7.972542611956545,
      "learning_rate": 5e-06,
      "loss": 0.1937,
      "step": 828
    },
    {
      "epoch": 1.436741767764298,
      "grad_norm": 8.427173766478871,
      "learning_rate": 5e-06,
      "loss": 0.2518,
      "step": 829
    },
    {
      "epoch": 1.438474870017331,
      "grad_norm": 8.030082885010385,
      "learning_rate": 5e-06,
      "loss": 0.2476,
      "step": 830
    },
    {
      "epoch": 1.440207972270364,
      "grad_norm": 8.434425076268536,
      "learning_rate": 5e-06,
      "loss": 0.3051,
      "step": 831
    },
    {
      "epoch": 1.441941074523397,
      "grad_norm": 10.88249176494796,
      "learning_rate": 5e-06,
      "loss": 0.3722,
      "step": 832
    },
    {
      "epoch": 1.4436741767764298,
      "grad_norm": 9.999856566354152,
      "learning_rate": 5e-06,
      "loss": 0.3361,
      "step": 833
    },
    {
      "epoch": 1.4454072790294628,
      "grad_norm": 6.501941390985206,
      "learning_rate": 5e-06,
      "loss": 0.2972,
      "step": 834
    },
    {
      "epoch": 1.4471403812824957,
      "grad_norm": 8.539252360223466,
      "learning_rate": 5e-06,
      "loss": 0.2208,
      "step": 835
    },
    {
      "epoch": 1.4488734835355286,
      "grad_norm": 7.2142795740773185,
      "learning_rate": 5e-06,
      "loss": 0.2692,
      "step": 836
    },
    {
      "epoch": 1.4506065857885615,
      "grad_norm": 10.532849832696968,
      "learning_rate": 5e-06,
      "loss": 0.4028,
      "step": 837
    },
    {
      "epoch": 1.4523396880415944,
      "grad_norm": 6.9951402961235285,
      "learning_rate": 5e-06,
      "loss": 0.2063,
      "step": 838
    },
    {
      "epoch": 1.4540727902946273,
      "grad_norm": 8.991427683607915,
      "learning_rate": 5e-06,
      "loss": 0.3029,
      "step": 839
    },
    {
      "epoch": 1.4558058925476602,
      "grad_norm": 6.250830328145195,
      "learning_rate": 5e-06,
      "loss": 0.2528,
      "step": 840
    },
    {
      "epoch": 1.4575389948006934,
      "grad_norm": 10.496281283214888,
      "learning_rate": 5e-06,
      "loss": 0.3157,
      "step": 841
    },
    {
      "epoch": 1.4592720970537263,
      "grad_norm": 7.655822492847662,
      "learning_rate": 5e-06,
      "loss": 0.2929,
      "step": 842
    },
    {
      "epoch": 1.4610051993067592,
      "grad_norm": 8.128150681187803,
      "learning_rate": 5e-06,
      "loss": 0.3124,
      "step": 843
    },
    {
      "epoch": 1.4627383015597921,
      "grad_norm": 13.328371714268437,
      "learning_rate": 5e-06,
      "loss": 0.3534,
      "step": 844
    },
    {
      "epoch": 1.464471403812825,
      "grad_norm": 7.690777583896878,
      "learning_rate": 5e-06,
      "loss": 0.288,
      "step": 845
    },
    {
      "epoch": 1.466204506065858,
      "grad_norm": 11.384240055043612,
      "learning_rate": 5e-06,
      "loss": 0.242,
      "step": 846
    },
    {
      "epoch": 1.4679376083188909,
      "grad_norm": 7.591957414449353,
      "learning_rate": 5e-06,
      "loss": 0.2209,
      "step": 847
    },
    {
      "epoch": 1.4696707105719238,
      "grad_norm": 6.079595628140821,
      "learning_rate": 5e-06,
      "loss": 0.2201,
      "step": 848
    },
    {
      "epoch": 1.4714038128249567,
      "grad_norm": 8.570395734412362,
      "learning_rate": 5e-06,
      "loss": 0.2887,
      "step": 849
    },
    {
      "epoch": 1.4731369150779896,
      "grad_norm": 8.19309934219369,
      "learning_rate": 5e-06,
      "loss": 0.2761,
      "step": 850
    },
    {
      "epoch": 1.4748700173310225,
      "grad_norm": 6.552695952756207,
      "learning_rate": 5e-06,
      "loss": 0.3067,
      "step": 851
    },
    {
      "epoch": 1.4766031195840554,
      "grad_norm": 9.929340587591632,
      "learning_rate": 5e-06,
      "loss": 0.2771,
      "step": 852
    },
    {
      "epoch": 1.4783362218370883,
      "grad_norm": 6.085583658178841,
      "learning_rate": 5e-06,
      "loss": 0.195,
      "step": 853
    },
    {
      "epoch": 1.4800693240901213,
      "grad_norm": 8.754796484602963,
      "learning_rate": 5e-06,
      "loss": 0.2688,
      "step": 854
    },
    {
      "epoch": 1.4818024263431542,
      "grad_norm": 9.051265514261969,
      "learning_rate": 5e-06,
      "loss": 0.2307,
      "step": 855
    },
    {
      "epoch": 1.483535528596187,
      "grad_norm": 5.867329561466726,
      "learning_rate": 5e-06,
      "loss": 0.1367,
      "step": 856
    },
    {
      "epoch": 1.48526863084922,
      "grad_norm": 6.475742854214952,
      "learning_rate": 5e-06,
      "loss": 0.3566,
      "step": 857
    },
    {
      "epoch": 1.487001733102253,
      "grad_norm": 7.066081261528999,
      "learning_rate": 5e-06,
      "loss": 0.1919,
      "step": 858
    },
    {
      "epoch": 1.4887348353552858,
      "grad_norm": 8.474951693866986,
      "learning_rate": 5e-06,
      "loss": 0.2126,
      "step": 859
    },
    {
      "epoch": 1.490467937608319,
      "grad_norm": 11.675819149319553,
      "learning_rate": 5e-06,
      "loss": 0.3321,
      "step": 860
    },
    {
      "epoch": 1.4922010398613519,
      "grad_norm": 6.702130721895754,
      "learning_rate": 5e-06,
      "loss": 0.2047,
      "step": 861
    },
    {
      "epoch": 1.4939341421143848,
      "grad_norm": 10.921084221193711,
      "learning_rate": 5e-06,
      "loss": 0.2381,
      "step": 862
    },
    {
      "epoch": 1.4956672443674177,
      "grad_norm": 10.155766965676833,
      "learning_rate": 5e-06,
      "loss": 0.3419,
      "step": 863
    },
    {
      "epoch": 1.4974003466204506,
      "grad_norm": 10.766328697293432,
      "learning_rate": 5e-06,
      "loss": 0.3888,
      "step": 864
    },
    {
      "epoch": 1.4991334488734835,
      "grad_norm": 10.818852140846912,
      "learning_rate": 5e-06,
      "loss": 0.2975,
      "step": 865
    },
    {
      "epoch": 1.5008665511265165,
      "grad_norm": 10.527102052536193,
      "learning_rate": 5e-06,
      "loss": 0.2742,
      "step": 866
    },
    {
      "epoch": 1.5025996533795494,
      "grad_norm": 6.02366485364541,
      "learning_rate": 5e-06,
      "loss": 0.175,
      "step": 867
    },
    {
      "epoch": 1.5043327556325823,
      "grad_norm": 14.402856522265871,
      "learning_rate": 5e-06,
      "loss": 0.4234,
      "step": 868
    },
    {
      "epoch": 1.5060658578856152,
      "grad_norm": 7.898361432079921,
      "learning_rate": 5e-06,
      "loss": 0.2553,
      "step": 869
    },
    {
      "epoch": 1.5077989601386483,
      "grad_norm": 9.812130672040366,
      "learning_rate": 5e-06,
      "loss": 0.3426,
      "step": 870
    },
    {
      "epoch": 1.5095320623916813,
      "grad_norm": 8.164133522418346,
      "learning_rate": 5e-06,
      "loss": 0.2627,
      "step": 871
    },
    {
      "epoch": 1.5112651646447142,
      "grad_norm": 7.140369101849324,
      "learning_rate": 5e-06,
      "loss": 0.2231,
      "step": 872
    },
    {
      "epoch": 1.512998266897747,
      "grad_norm": 7.777667710872711,
      "learning_rate": 5e-06,
      "loss": 0.2139,
      "step": 873
    },
    {
      "epoch": 1.51473136915078,
      "grad_norm": 5.424552146602606,
      "learning_rate": 5e-06,
      "loss": 0.1391,
      "step": 874
    },
    {
      "epoch": 1.516464471403813,
      "grad_norm": 6.079290989142078,
      "learning_rate": 5e-06,
      "loss": 0.2106,
      "step": 875
    },
    {
      "epoch": 1.5181975736568458,
      "grad_norm": 9.782776734756336,
      "learning_rate": 5e-06,
      "loss": 0.2003,
      "step": 876
    },
    {
      "epoch": 1.5199306759098787,
      "grad_norm": 9.425501592023124,
      "learning_rate": 5e-06,
      "loss": 0.3939,
      "step": 877
    },
    {
      "epoch": 1.5216637781629117,
      "grad_norm": 10.771605315787502,
      "learning_rate": 5e-06,
      "loss": 0.383,
      "step": 878
    },
    {
      "epoch": 1.5233968804159446,
      "grad_norm": 10.49196971630361,
      "learning_rate": 5e-06,
      "loss": 0.4436,
      "step": 879
    },
    {
      "epoch": 1.5251299826689775,
      "grad_norm": 10.798199143205691,
      "learning_rate": 5e-06,
      "loss": 0.2977,
      "step": 880
    },
    {
      "epoch": 1.5268630849220104,
      "grad_norm": 6.4510610018288,
      "learning_rate": 5e-06,
      "loss": 0.1878,
      "step": 881
    },
    {
      "epoch": 1.5285961871750433,
      "grad_norm": 6.916179992222184,
      "learning_rate": 5e-06,
      "loss": 0.2903,
      "step": 882
    },
    {
      "epoch": 1.5303292894280762,
      "grad_norm": 7.4725919283039355,
      "learning_rate": 5e-06,
      "loss": 0.299,
      "step": 883
    },
    {
      "epoch": 1.5320623916811091,
      "grad_norm": 10.540632194654563,
      "learning_rate": 5e-06,
      "loss": 0.3,
      "step": 884
    },
    {
      "epoch": 1.533795493934142,
      "grad_norm": 10.407238979140688,
      "learning_rate": 5e-06,
      "loss": 0.3205,
      "step": 885
    },
    {
      "epoch": 1.535528596187175,
      "grad_norm": 7.9322309652415885,
      "learning_rate": 5e-06,
      "loss": 0.2159,
      "step": 886
    },
    {
      "epoch": 1.5372616984402079,
      "grad_norm": 7.564387566685964,
      "learning_rate": 5e-06,
      "loss": 0.3666,
      "step": 887
    },
    {
      "epoch": 1.5389948006932408,
      "grad_norm": 7.7286947951179945,
      "learning_rate": 5e-06,
      "loss": 0.2944,
      "step": 888
    },
    {
      "epoch": 1.5407279029462737,
      "grad_norm": 7.3371063698717425,
      "learning_rate": 5e-06,
      "loss": 0.2606,
      "step": 889
    },
    {
      "epoch": 1.5424610051993066,
      "grad_norm": 6.796130189974441,
      "learning_rate": 5e-06,
      "loss": 0.2457,
      "step": 890
    },
    {
      "epoch": 1.5441941074523395,
      "grad_norm": 7.869963958153603,
      "learning_rate": 5e-06,
      "loss": 0.3522,
      "step": 891
    },
    {
      "epoch": 1.5459272097053725,
      "grad_norm": 8.462838562541798,
      "learning_rate": 5e-06,
      "loss": 0.2625,
      "step": 892
    },
    {
      "epoch": 1.5476603119584056,
      "grad_norm": 5.937117393614231,
      "learning_rate": 5e-06,
      "loss": 0.2804,
      "step": 893
    },
    {
      "epoch": 1.5493934142114385,
      "grad_norm": 4.002034861826958,
      "learning_rate": 5e-06,
      "loss": 0.1669,
      "step": 894
    },
    {
      "epoch": 1.5511265164644714,
      "grad_norm": 6.311488910942427,
      "learning_rate": 5e-06,
      "loss": 0.2915,
      "step": 895
    },
    {
      "epoch": 1.5528596187175043,
      "grad_norm": 9.004841667787062,
      "learning_rate": 5e-06,
      "loss": 0.227,
      "step": 896
    },
    {
      "epoch": 1.5545927209705372,
      "grad_norm": 10.049282325437357,
      "learning_rate": 5e-06,
      "loss": 0.2847,
      "step": 897
    },
    {
      "epoch": 1.5563258232235702,
      "grad_norm": 11.450150961068156,
      "learning_rate": 5e-06,
      "loss": 0.25,
      "step": 898
    },
    {
      "epoch": 1.558058925476603,
      "grad_norm": 9.239360980111167,
      "learning_rate": 5e-06,
      "loss": 0.265,
      "step": 899
    },
    {
      "epoch": 1.5597920277296362,
      "grad_norm": 7.095307435464637,
      "learning_rate": 5e-06,
      "loss": 0.2435,
      "step": 900
    },
    {
      "epoch": 1.5615251299826691,
      "grad_norm": 7.702206818059618,
      "learning_rate": 5e-06,
      "loss": 0.2323,
      "step": 901
    },
    {
      "epoch": 1.563258232235702,
      "grad_norm": 14.808431460961163,
      "learning_rate": 5e-06,
      "loss": 0.4353,
      "step": 902
    },
    {
      "epoch": 1.564991334488735,
      "grad_norm": 12.263791786770428,
      "learning_rate": 5e-06,
      "loss": 0.3965,
      "step": 903
    },
    {
      "epoch": 1.5667244367417679,
      "grad_norm": 13.673521106260793,
      "learning_rate": 5e-06,
      "loss": 0.4162,
      "step": 904
    },
    {
      "epoch": 1.5684575389948008,
      "grad_norm": 14.335604879485977,
      "learning_rate": 5e-06,
      "loss": 0.1785,
      "step": 905
    },
    {
      "epoch": 1.5701906412478337,
      "grad_norm": 11.011043553594734,
      "learning_rate": 5e-06,
      "loss": 0.3112,
      "step": 906
    },
    {
      "epoch": 1.5719237435008666,
      "grad_norm": 14.216898166669315,
      "learning_rate": 5e-06,
      "loss": 0.4534,
      "step": 907
    },
    {
      "epoch": 1.5736568457538995,
      "grad_norm": 7.101453455055268,
      "learning_rate": 5e-06,
      "loss": 0.3579,
      "step": 908
    },
    {
      "epoch": 1.5753899480069324,
      "grad_norm": 7.363114688517338,
      "learning_rate": 5e-06,
      "loss": 0.3443,
      "step": 909
    },
    {
      "epoch": 1.5771230502599654,
      "grad_norm": 8.54038071295232,
      "learning_rate": 5e-06,
      "loss": 0.2159,
      "step": 910
    },
    {
      "epoch": 1.5788561525129983,
      "grad_norm": 7.652313590659685,
      "learning_rate": 5e-06,
      "loss": 0.1866,
      "step": 911
    },
    {
      "epoch": 1.5805892547660312,
      "grad_norm": 13.177318529686247,
      "learning_rate": 5e-06,
      "loss": 0.4013,
      "step": 912
    },
    {
      "epoch": 1.582322357019064,
      "grad_norm": 6.954062820795925,
      "learning_rate": 5e-06,
      "loss": 0.1788,
      "step": 913
    },
    {
      "epoch": 1.584055459272097,
      "grad_norm": 5.703895830291983,
      "learning_rate": 5e-06,
      "loss": 0.1616,
      "step": 914
    },
    {
      "epoch": 1.58578856152513,
      "grad_norm": 9.174768046843413,
      "learning_rate": 5e-06,
      "loss": 0.2787,
      "step": 915
    },
    {
      "epoch": 1.5875216637781628,
      "grad_norm": 8.490434986918505,
      "learning_rate": 5e-06,
      "loss": 0.2162,
      "step": 916
    },
    {
      "epoch": 1.5892547660311958,
      "grad_norm": 6.098504505380679,
      "learning_rate": 5e-06,
      "loss": 0.1612,
      "step": 917
    },
    {
      "epoch": 1.5909878682842287,
      "grad_norm": 10.478949744909919,
      "learning_rate": 5e-06,
      "loss": 0.3619,
      "step": 918
    },
    {
      "epoch": 1.5927209705372616,
      "grad_norm": 6.637870809825953,
      "learning_rate": 5e-06,
      "loss": 0.259,
      "step": 919
    },
    {
      "epoch": 1.5944540727902945,
      "grad_norm": 10.068488382721027,
      "learning_rate": 5e-06,
      "loss": 0.3119,
      "step": 920
    },
    {
      "epoch": 1.5961871750433274,
      "grad_norm": 13.280238932103217,
      "learning_rate": 5e-06,
      "loss": 0.3609,
      "step": 921
    },
    {
      "epoch": 1.5979202772963603,
      "grad_norm": 7.696685513616092,
      "learning_rate": 5e-06,
      "loss": 0.2429,
      "step": 922
    },
    {
      "epoch": 1.5996533795493935,
      "grad_norm": 6.1647738395597225,
      "learning_rate": 5e-06,
      "loss": 0.151,
      "step": 923
    },
    {
      "epoch": 1.6013864818024264,
      "grad_norm": 10.963428391017151,
      "learning_rate": 5e-06,
      "loss": 0.2352,
      "step": 924
    },
    {
      "epoch": 1.6031195840554593,
      "grad_norm": 9.224903600338969,
      "learning_rate": 5e-06,
      "loss": 0.3887,
      "step": 925
    },
    {
      "epoch": 1.6048526863084922,
      "grad_norm": 7.057482532205292,
      "learning_rate": 5e-06,
      "loss": 0.2371,
      "step": 926
    },
    {
      "epoch": 1.6065857885615251,
      "grad_norm": 8.142092991589534,
      "learning_rate": 5e-06,
      "loss": 0.2486,
      "step": 927
    },
    {
      "epoch": 1.608318890814558,
      "grad_norm": 6.221108760813236,
      "learning_rate": 5e-06,
      "loss": 0.2482,
      "step": 928
    },
    {
      "epoch": 1.610051993067591,
      "grad_norm": 7.552172458166631,
      "learning_rate": 5e-06,
      "loss": 0.2649,
      "step": 929
    },
    {
      "epoch": 1.6117850953206239,
      "grad_norm": 7.43757719112057,
      "learning_rate": 5e-06,
      "loss": 0.2126,
      "step": 930
    },
    {
      "epoch": 1.613518197573657,
      "grad_norm": 11.482067556246607,
      "learning_rate": 5e-06,
      "loss": 0.2283,
      "step": 931
    },
    {
      "epoch": 1.61525129982669,
      "grad_norm": 6.068159470275106,
      "learning_rate": 5e-06,
      "loss": 0.1829,
      "step": 932
    },
    {
      "epoch": 1.6169844020797228,
      "grad_norm": 12.007898274539125,
      "learning_rate": 5e-06,
      "loss": 0.2939,
      "step": 933
    },
    {
      "epoch": 1.6187175043327557,
      "grad_norm": 9.707966897185809,
      "learning_rate": 5e-06,
      "loss": 0.2757,
      "step": 934
    },
    {
      "epoch": 1.6204506065857887,
      "grad_norm": 5.271131584966852,
      "learning_rate": 5e-06,
      "loss": 0.165,
      "step": 935
    },
    {
      "epoch": 1.6221837088388216,
      "grad_norm": 9.47642281752476,
      "learning_rate": 5e-06,
      "loss": 0.2315,
      "step": 936
    },
    {
      "epoch": 1.6239168110918545,
      "grad_norm": 8.60781563762082,
      "learning_rate": 5e-06,
      "loss": 0.2609,
      "step": 937
    },
    {
      "epoch": 1.6256499133448874,
      "grad_norm": 13.410417384764493,
      "learning_rate": 5e-06,
      "loss": 0.3933,
      "step": 938
    },
    {
      "epoch": 1.6273830155979203,
      "grad_norm": 13.12872959007893,
      "learning_rate": 5e-06,
      "loss": 0.2991,
      "step": 939
    },
    {
      "epoch": 1.6291161178509532,
      "grad_norm": 8.575938573247209,
      "learning_rate": 5e-06,
      "loss": 0.2309,
      "step": 940
    },
    {
      "epoch": 1.6308492201039861,
      "grad_norm": 8.939108110353837,
      "learning_rate": 5e-06,
      "loss": 0.3936,
      "step": 941
    },
    {
      "epoch": 1.632582322357019,
      "grad_norm": 8.812944549818953,
      "learning_rate": 5e-06,
      "loss": 0.3373,
      "step": 942
    },
    {
      "epoch": 1.634315424610052,
      "grad_norm": 5.744546045175086,
      "learning_rate": 5e-06,
      "loss": 0.19,
      "step": 943
    },
    {
      "epoch": 1.6360485268630849,
      "grad_norm": 7.834587186200293,
      "learning_rate": 5e-06,
      "loss": 0.2402,
      "step": 944
    },
    {
      "epoch": 1.6377816291161178,
      "grad_norm": 10.89859605899443,
      "learning_rate": 5e-06,
      "loss": 0.4358,
      "step": 945
    },
    {
      "epoch": 1.6395147313691507,
      "grad_norm": 6.8220797037542,
      "learning_rate": 5e-06,
      "loss": 0.3321,
      "step": 946
    },
    {
      "epoch": 1.6412478336221836,
      "grad_norm": 9.167468596551068,
      "learning_rate": 5e-06,
      "loss": 0.182,
      "step": 947
    },
    {
      "epoch": 1.6429809358752165,
      "grad_norm": 6.356921723317562,
      "learning_rate": 5e-06,
      "loss": 0.2311,
      "step": 948
    },
    {
      "epoch": 1.6447140381282495,
      "grad_norm": 10.620789535110895,
      "learning_rate": 5e-06,
      "loss": 0.3007,
      "step": 949
    },
    {
      "epoch": 1.6464471403812824,
      "grad_norm": 8.748886909530933,
      "learning_rate": 5e-06,
      "loss": 0.2711,
      "step": 950
    },
    {
      "epoch": 1.6481802426343153,
      "grad_norm": 12.075808442341772,
      "learning_rate": 5e-06,
      "loss": 0.3433,
      "step": 951
    },
    {
      "epoch": 1.6499133448873482,
      "grad_norm": 8.016057588381829,
      "learning_rate": 5e-06,
      "loss": 0.1657,
      "step": 952
    },
    {
      "epoch": 1.6516464471403813,
      "grad_norm": 8.51414076443084,
      "learning_rate": 5e-06,
      "loss": 0.222,
      "step": 953
    },
    {
      "epoch": 1.6533795493934142,
      "grad_norm": 6.20803907409641,
      "learning_rate": 5e-06,
      "loss": 0.2195,
      "step": 954
    },
    {
      "epoch": 1.6551126516464472,
      "grad_norm": 4.909134326768502,
      "learning_rate": 5e-06,
      "loss": 0.1745,
      "step": 955
    },
    {
      "epoch": 1.65684575389948,
      "grad_norm": 6.813986239917366,
      "learning_rate": 5e-06,
      "loss": 0.2284,
      "step": 956
    },
    {
      "epoch": 1.658578856152513,
      "grad_norm": 6.835982979759425,
      "learning_rate": 5e-06,
      "loss": 0.1929,
      "step": 957
    },
    {
      "epoch": 1.660311958405546,
      "grad_norm": 9.734855823673756,
      "learning_rate": 5e-06,
      "loss": 0.1875,
      "step": 958
    },
    {
      "epoch": 1.6620450606585788,
      "grad_norm": 8.74501156980121,
      "learning_rate": 5e-06,
      "loss": 0.2537,
      "step": 959
    },
    {
      "epoch": 1.6637781629116117,
      "grad_norm": 7.393515926728573,
      "learning_rate": 5e-06,
      "loss": 0.2791,
      "step": 960
    },
    {
      "epoch": 1.6655112651646449,
      "grad_norm": 12.384364428577054,
      "learning_rate": 5e-06,
      "loss": 0.2549,
      "step": 961
    },
    {
      "epoch": 1.6672443674176778,
      "grad_norm": 9.472912770366909,
      "learning_rate": 5e-06,
      "loss": 0.4049,
      "step": 962
    },
    {
      "epoch": 1.6689774696707107,
      "grad_norm": 11.311766542715356,
      "learning_rate": 5e-06,
      "loss": 0.4008,
      "step": 963
    },
    {
      "epoch": 1.6707105719237436,
      "grad_norm": 9.411454190012462,
      "learning_rate": 5e-06,
      "loss": 0.2894,
      "step": 964
    },
    {
      "epoch": 1.6724436741767765,
      "grad_norm": 10.328700062588233,
      "learning_rate": 5e-06,
      "loss": 0.3406,
      "step": 965
    },
    {
      "epoch": 1.6741767764298094,
      "grad_norm": 6.775382179657464,
      "learning_rate": 5e-06,
      "loss": 0.167,
      "step": 966
    },
    {
      "epoch": 1.6759098786828424,
      "grad_norm": 10.21290276432392,
      "learning_rate": 5e-06,
      "loss": 0.3592,
      "step": 967
    },
    {
      "epoch": 1.6776429809358753,
      "grad_norm": 6.535885274825714,
      "learning_rate": 5e-06,
      "loss": 0.1753,
      "step": 968
    },
    {
      "epoch": 1.6793760831889082,
      "grad_norm": 7.289387490242932,
      "learning_rate": 5e-06,
      "loss": 0.3269,
      "step": 969
    },
    {
      "epoch": 1.681109185441941,
      "grad_norm": 8.40984962283577,
      "learning_rate": 5e-06,
      "loss": 0.28,
      "step": 970
    },
    {
      "epoch": 1.682842287694974,
      "grad_norm": 6.498520902927162,
      "learning_rate": 5e-06,
      "loss": 0.2347,
      "step": 971
    },
    {
      "epoch": 1.684575389948007,
      "grad_norm": 8.672619267047981,
      "learning_rate": 5e-06,
      "loss": 0.2788,
      "step": 972
    },
    {
      "epoch": 1.6863084922010398,
      "grad_norm": 5.23713233403821,
      "learning_rate": 5e-06,
      "loss": 0.1983,
      "step": 973
    },
    {
      "epoch": 1.6880415944540728,
      "grad_norm": 6.602759402869207,
      "learning_rate": 5e-06,
      "loss": 0.308,
      "step": 974
    },
    {
      "epoch": 1.6897746967071057,
      "grad_norm": 8.054244672144591,
      "learning_rate": 5e-06,
      "loss": 0.3171,
      "step": 975
    },
    {
      "epoch": 1.6915077989601386,
      "grad_norm": 10.441811042471496,
      "learning_rate": 5e-06,
      "loss": 0.3031,
      "step": 976
    },
    {
      "epoch": 1.6932409012131715,
      "grad_norm": 7.658262151886139,
      "learning_rate": 5e-06,
      "loss": 0.2704,
      "step": 977
    },
    {
      "epoch": 1.6949740034662044,
      "grad_norm": 5.650864708820896,
      "learning_rate": 5e-06,
      "loss": 0.1445,
      "step": 978
    },
    {
      "epoch": 1.6967071057192373,
      "grad_norm": 5.88535062589699,
      "learning_rate": 5e-06,
      "loss": 0.1504,
      "step": 979
    },
    {
      "epoch": 1.6984402079722702,
      "grad_norm": 5.729112585563876,
      "learning_rate": 5e-06,
      "loss": 0.1974,
      "step": 980
    },
    {
      "epoch": 1.7001733102253032,
      "grad_norm": 7.341784404917582,
      "learning_rate": 5e-06,
      "loss": 0.2785,
      "step": 981
    },
    {
      "epoch": 1.701906412478336,
      "grad_norm": 8.016695244857807,
      "learning_rate": 5e-06,
      "loss": 0.2388,
      "step": 982
    },
    {
      "epoch": 1.703639514731369,
      "grad_norm": 7.597736332723833,
      "learning_rate": 5e-06,
      "loss": 0.2237,
      "step": 983
    },
    {
      "epoch": 1.7053726169844021,
      "grad_norm": 8.054992492407077,
      "learning_rate": 5e-06,
      "loss": 0.1863,
      "step": 984
    },
    {
      "epoch": 1.707105719237435,
      "grad_norm": 5.291087722146295,
      "learning_rate": 5e-06,
      "loss": 0.2415,
      "step": 985
    },
    {
      "epoch": 1.708838821490468,
      "grad_norm": 9.095664330830411,
      "learning_rate": 5e-06,
      "loss": 0.2708,
      "step": 986
    },
    {
      "epoch": 1.7105719237435009,
      "grad_norm": 6.818448051109842,
      "learning_rate": 5e-06,
      "loss": 0.1806,
      "step": 987
    },
    {
      "epoch": 1.7123050259965338,
      "grad_norm": 9.796443558582082,
      "learning_rate": 5e-06,
      "loss": 0.2443,
      "step": 988
    },
    {
      "epoch": 1.7140381282495667,
      "grad_norm": 6.318354846887868,
      "learning_rate": 5e-06,
      "loss": 0.1914,
      "step": 989
    },
    {
      "epoch": 1.7157712305025996,
      "grad_norm": 8.995710410308504,
      "learning_rate": 5e-06,
      "loss": 0.3058,
      "step": 990
    },
    {
      "epoch": 1.7175043327556327,
      "grad_norm": 8.14788178609003,
      "learning_rate": 5e-06,
      "loss": 0.2592,
      "step": 991
    },
    {
      "epoch": 1.7192374350086657,
      "grad_norm": 8.064769809793763,
      "learning_rate": 5e-06,
      "loss": 0.2019,
      "step": 992
    },
    {
      "epoch": 1.7209705372616986,
      "grad_norm": 9.207390462206714,
      "learning_rate": 5e-06,
      "loss": 0.3095,
      "step": 993
    },
    {
      "epoch": 1.7227036395147315,
      "grad_norm": 6.6535530178417055,
      "learning_rate": 5e-06,
      "loss": 0.245,
      "step": 994
    },
    {
      "epoch": 1.7244367417677644,
      "grad_norm": 7.039438862111681,
      "learning_rate": 5e-06,
      "loss": 0.2224,
      "step": 995
    },
    {
      "epoch": 1.7261698440207973,
      "grad_norm": 7.5402383312330965,
      "learning_rate": 5e-06,
      "loss": 0.2585,
      "step": 996
    },
    {
      "epoch": 1.7279029462738302,
      "grad_norm": 12.992257967337462,
      "learning_rate": 5e-06,
      "loss": 0.4281,
      "step": 997
    },
    {
      "epoch": 1.7296360485268631,
      "grad_norm": 9.063739981171627,
      "learning_rate": 5e-06,
      "loss": 0.3143,
      "step": 998
    },
    {
      "epoch": 1.731369150779896,
      "grad_norm": 7.502134400878703,
      "learning_rate": 5e-06,
      "loss": 0.3558,
      "step": 999
    },
    {
      "epoch": 1.733102253032929,
      "grad_norm": 8.254479319315532,
      "learning_rate": 5e-06,
      "loss": 0.3121,
      "step": 1000
    },
    {
      "epoch": 1.7348353552859619,
      "grad_norm": 11.495153981721018,
      "learning_rate": 5e-06,
      "loss": 0.3008,
      "step": 1001
    },
    {
      "epoch": 1.7365684575389948,
      "grad_norm": 10.415320673129614,
      "learning_rate": 5e-06,
      "loss": 0.3056,
      "step": 1002
    },
    {
      "epoch": 1.7383015597920277,
      "grad_norm": 6.3208698644653225,
      "learning_rate": 5e-06,
      "loss": 0.1187,
      "step": 1003
    },
    {
      "epoch": 1.7400346620450606,
      "grad_norm": 9.190281038355947,
      "learning_rate": 5e-06,
      "loss": 0.3449,
      "step": 1004
    },
    {
      "epoch": 1.7417677642980935,
      "grad_norm": 5.8688731610229725,
      "learning_rate": 5e-06,
      "loss": 0.3211,
      "step": 1005
    },
    {
      "epoch": 1.7435008665511265,
      "grad_norm": 8.231322908187217,
      "learning_rate": 5e-06,
      "loss": 0.2877,
      "step": 1006
    },
    {
      "epoch": 1.7452339688041594,
      "grad_norm": 5.658831097034583,
      "learning_rate": 5e-06,
      "loss": 0.2303,
      "step": 1007
    },
    {
      "epoch": 1.7469670710571923,
      "grad_norm": 8.78157884518269,
      "learning_rate": 5e-06,
      "loss": 0.2678,
      "step": 1008
    },
    {
      "epoch": 1.7487001733102252,
      "grad_norm": 6.5369903307430794,
      "learning_rate": 5e-06,
      "loss": 0.2485,
      "step": 1009
    },
    {
      "epoch": 1.7504332755632581,
      "grad_norm": 7.5021392314523,
      "learning_rate": 5e-06,
      "loss": 0.313,
      "step": 1010
    },
    {
      "epoch": 1.752166377816291,
      "grad_norm": 5.0538218020750305,
      "learning_rate": 5e-06,
      "loss": 0.1622,
      "step": 1011
    },
    {
      "epoch": 1.753899480069324,
      "grad_norm": 7.285245789496666,
      "learning_rate": 5e-06,
      "loss": 0.1814,
      "step": 1012
    },
    {
      "epoch": 1.7556325823223569,
      "grad_norm": 8.681966872033655,
      "learning_rate": 5e-06,
      "loss": 0.3473,
      "step": 1013
    },
    {
      "epoch": 1.75736568457539,
      "grad_norm": 15.320616272166873,
      "learning_rate": 5e-06,
      "loss": 0.4882,
      "step": 1014
    },
    {
      "epoch": 1.759098786828423,
      "grad_norm": 9.721006272586195,
      "learning_rate": 5e-06,
      "loss": 0.279,
      "step": 1015
    },
    {
      "epoch": 1.7608318890814558,
      "grad_norm": 6.4172492155047625,
      "learning_rate": 5e-06,
      "loss": 0.2431,
      "step": 1016
    },
    {
      "epoch": 1.7625649913344887,
      "grad_norm": 8.961052013353966,
      "learning_rate": 5e-06,
      "loss": 0.2236,
      "step": 1017
    },
    {
      "epoch": 1.7642980935875217,
      "grad_norm": 7.2129220928267,
      "learning_rate": 5e-06,
      "loss": 0.2781,
      "step": 1018
    },
    {
      "epoch": 1.7660311958405546,
      "grad_norm": 9.086815356210948,
      "learning_rate": 5e-06,
      "loss": 0.3645,
      "step": 1019
    },
    {
      "epoch": 1.7677642980935875,
      "grad_norm": 8.446115573808637,
      "learning_rate": 5e-06,
      "loss": 0.2366,
      "step": 1020
    },
    {
      "epoch": 1.7694974003466204,
      "grad_norm": 8.72108171245698,
      "learning_rate": 5e-06,
      "loss": 0.3066,
      "step": 1021
    },
    {
      "epoch": 1.7712305025996535,
      "grad_norm": 5.764959364100198,
      "learning_rate": 5e-06,
      "loss": 0.3203,
      "step": 1022
    },
    {
      "epoch": 1.7729636048526864,
      "grad_norm": 7.92156584548383,
      "learning_rate": 5e-06,
      "loss": 0.2957,
      "step": 1023
    },
    {
      "epoch": 1.7746967071057194,
      "grad_norm": 5.345794286706713,
      "learning_rate": 5e-06,
      "loss": 0.3256,
      "step": 1024
    },
    {
      "epoch": 1.7764298093587523,
      "grad_norm": 8.049068174274653,
      "learning_rate": 5e-06,
      "loss": 0.2592,
      "step": 1025
    },
    {
      "epoch": 1.7781629116117852,
      "grad_norm": 8.839167078056,
      "learning_rate": 5e-06,
      "loss": 0.2545,
      "step": 1026
    },
    {
      "epoch": 1.779896013864818,
      "grad_norm": 8.033294536238357,
      "learning_rate": 5e-06,
      "loss": 0.2532,
      "step": 1027
    },
    {
      "epoch": 1.781629116117851,
      "grad_norm": 8.774993513925443,
      "learning_rate": 5e-06,
      "loss": 0.232,
      "step": 1028
    },
    {
      "epoch": 1.783362218370884,
      "grad_norm": 5.265002477052981,
      "learning_rate": 5e-06,
      "loss": 0.1405,
      "step": 1029
    },
    {
      "epoch": 1.7850953206239168,
      "grad_norm": 8.63863897694674,
      "learning_rate": 5e-06,
      "loss": 0.2817,
      "step": 1030
    },
    {
      "epoch": 1.7868284228769498,
      "grad_norm": 10.532809993764536,
      "learning_rate": 5e-06,
      "loss": 0.4216,
      "step": 1031
    },
    {
      "epoch": 1.7885615251299827,
      "grad_norm": 8.363408963877728,
      "learning_rate": 5e-06,
      "loss": 0.2439,
      "step": 1032
    },
    {
      "epoch": 1.7902946273830156,
      "grad_norm": 8.205820334721121,
      "learning_rate": 5e-06,
      "loss": 0.2723,
      "step": 1033
    },
    {
      "epoch": 1.7920277296360485,
      "grad_norm": 12.284209770774138,
      "learning_rate": 5e-06,
      "loss": 0.289,
      "step": 1034
    },
    {
      "epoch": 1.7937608318890814,
      "grad_norm": 5.429293699186063,
      "learning_rate": 5e-06,
      "loss": 0.2859,
      "step": 1035
    },
    {
      "epoch": 1.7954939341421143,
      "grad_norm": 7.049970737694147,
      "learning_rate": 5e-06,
      "loss": 0.1365,
      "step": 1036
    },
    {
      "epoch": 1.7972270363951472,
      "grad_norm": 8.411928209028394,
      "learning_rate": 5e-06,
      "loss": 0.31,
      "step": 1037
    },
    {
      "epoch": 1.7989601386481802,
      "grad_norm": 7.29933036972814,
      "learning_rate": 5e-06,
      "loss": 0.2234,
      "step": 1038
    },
    {
      "epoch": 1.800693240901213,
      "grad_norm": 5.712661921170019,
      "learning_rate": 5e-06,
      "loss": 0.0956,
      "step": 1039
    },
    {
      "epoch": 1.802426343154246,
      "grad_norm": 9.100820617564437,
      "learning_rate": 5e-06,
      "loss": 0.3103,
      "step": 1040
    },
    {
      "epoch": 1.804159445407279,
      "grad_norm": 9.264239099819953,
      "learning_rate": 5e-06,
      "loss": 0.2504,
      "step": 1041
    },
    {
      "epoch": 1.8058925476603118,
      "grad_norm": 11.269402959083468,
      "learning_rate": 5e-06,
      "loss": 0.3482,
      "step": 1042
    },
    {
      "epoch": 1.8076256499133447,
      "grad_norm": 7.613424801830958,
      "learning_rate": 5e-06,
      "loss": 0.2315,
      "step": 1043
    },
    {
      "epoch": 1.8093587521663779,
      "grad_norm": 12.240251671690746,
      "learning_rate": 5e-06,
      "loss": 0.2166,
      "step": 1044
    },
    {
      "epoch": 1.8110918544194108,
      "grad_norm": 10.15426212516895,
      "learning_rate": 5e-06,
      "loss": 0.4241,
      "step": 1045
    },
    {
      "epoch": 1.8128249566724437,
      "grad_norm": 16.02634264101656,
      "learning_rate": 5e-06,
      "loss": 0.4542,
      "step": 1046
    },
    {
      "epoch": 1.8145580589254766,
      "grad_norm": 8.291378866085939,
      "learning_rate": 5e-06,
      "loss": 0.2227,
      "step": 1047
    },
    {
      "epoch": 1.8162911611785095,
      "grad_norm": 11.46976044488663,
      "learning_rate": 5e-06,
      "loss": 0.2954,
      "step": 1048
    },
    {
      "epoch": 1.8180242634315424,
      "grad_norm": 5.043593340666032,
      "learning_rate": 5e-06,
      "loss": 0.1547,
      "step": 1049
    },
    {
      "epoch": 1.8197573656845754,
      "grad_norm": 12.992777064945546,
      "learning_rate": 5e-06,
      "loss": 0.3679,
      "step": 1050
    },
    {
      "epoch": 1.8214904679376083,
      "grad_norm": 10.53909109463782,
      "learning_rate": 5e-06,
      "loss": 0.3143,
      "step": 1051
    },
    {
      "epoch": 1.8232235701906414,
      "grad_norm": 9.66149128019708,
      "learning_rate": 5e-06,
      "loss": 0.2678,
      "step": 1052
    },
    {
      "epoch": 1.8249566724436743,
      "grad_norm": 7.338410729421692,
      "learning_rate": 5e-06,
      "loss": 0.2649,
      "step": 1053
    },
    {
      "epoch": 1.8266897746967072,
      "grad_norm": 11.263140315780019,
      "learning_rate": 5e-06,
      "loss": 0.3692,
      "step": 1054
    },
    {
      "epoch": 1.8284228769497402,
      "grad_norm": 6.127686514721937,
      "learning_rate": 5e-06,
      "loss": 0.2441,
      "step": 1055
    },
    {
      "epoch": 1.830155979202773,
      "grad_norm": 6.915468165726155,
      "learning_rate": 5e-06,
      "loss": 0.2373,
      "step": 1056
    },
    {
      "epoch": 1.831889081455806,
      "grad_norm": 4.946037926281287,
      "learning_rate": 5e-06,
      "loss": 0.2127,
      "step": 1057
    },
    {
      "epoch": 1.833622183708839,
      "grad_norm": 6.873972035630554,
      "learning_rate": 5e-06,
      "loss": 0.308,
      "step": 1058
    },
    {
      "epoch": 1.8353552859618718,
      "grad_norm": 8.90987111145028,
      "learning_rate": 5e-06,
      "loss": 0.3327,
      "step": 1059
    },
    {
      "epoch": 1.8370883882149047,
      "grad_norm": 6.534820310870172,
      "learning_rate": 5e-06,
      "loss": 0.206,
      "step": 1060
    },
    {
      "epoch": 1.8388214904679376,
      "grad_norm": 10.697702278779783,
      "learning_rate": 5e-06,
      "loss": 0.3037,
      "step": 1061
    },
    {
      "epoch": 1.8405545927209706,
      "grad_norm": 5.136873772316189,
      "learning_rate": 5e-06,
      "loss": 0.192,
      "step": 1062
    },
    {
      "epoch": 1.8422876949740035,
      "grad_norm": 8.458907936259923,
      "learning_rate": 5e-06,
      "loss": 0.386,
      "step": 1063
    },
    {
      "epoch": 1.8440207972270364,
      "grad_norm": 10.350463163912982,
      "learning_rate": 5e-06,
      "loss": 0.3479,
      "step": 1064
    },
    {
      "epoch": 1.8457538994800693,
      "grad_norm": 8.437000853291263,
      "learning_rate": 5e-06,
      "loss": 0.2613,
      "step": 1065
    },
    {
      "epoch": 1.8474870017331022,
      "grad_norm": 9.422053953783529,
      "learning_rate": 5e-06,
      "loss": 0.3406,
      "step": 1066
    },
    {
      "epoch": 1.8492201039861351,
      "grad_norm": 6.121244563793022,
      "learning_rate": 5e-06,
      "loss": 0.2385,
      "step": 1067
    },
    {
      "epoch": 1.850953206239168,
      "grad_norm": 9.644200186070584,
      "learning_rate": 5e-06,
      "loss": 0.2911,
      "step": 1068
    },
    {
      "epoch": 1.852686308492201,
      "grad_norm": 8.042123521335759,
      "learning_rate": 5e-06,
      "loss": 0.2975,
      "step": 1069
    },
    {
      "epoch": 1.8544194107452339,
      "grad_norm": 6.770637892812312,
      "learning_rate": 5e-06,
      "loss": 0.2093,
      "step": 1070
    },
    {
      "epoch": 1.8561525129982668,
      "grad_norm": 7.285115407057092,
      "learning_rate": 5e-06,
      "loss": 0.2121,
      "step": 1071
    },
    {
      "epoch": 1.8578856152512997,
      "grad_norm": 6.241488039171046,
      "learning_rate": 5e-06,
      "loss": 0.2516,
      "step": 1072
    },
    {
      "epoch": 1.8596187175043326,
      "grad_norm": 7.8811115588523535,
      "learning_rate": 5e-06,
      "loss": 0.253,
      "step": 1073
    },
    {
      "epoch": 1.8613518197573655,
      "grad_norm": 9.665850857039299,
      "learning_rate": 5e-06,
      "loss": 0.2938,
      "step": 1074
    },
    {
      "epoch": 1.8630849220103987,
      "grad_norm": 8.426865947797463,
      "learning_rate": 5e-06,
      "loss": 0.2819,
      "step": 1075
    },
    {
      "epoch": 1.8648180242634316,
      "grad_norm": 6.0961696418126206,
      "learning_rate": 5e-06,
      "loss": 0.2244,
      "step": 1076
    },
    {
      "epoch": 1.8665511265164645,
      "grad_norm": 6.213605583567767,
      "learning_rate": 5e-06,
      "loss": 0.2356,
      "step": 1077
    },
    {
      "epoch": 1.8682842287694974,
      "grad_norm": 8.79802364350871,
      "learning_rate": 5e-06,
      "loss": 0.338,
      "step": 1078
    },
    {
      "epoch": 1.8700173310225303,
      "grad_norm": 6.573998831177582,
      "learning_rate": 5e-06,
      "loss": 0.3038,
      "step": 1079
    },
    {
      "epoch": 1.8717504332755632,
      "grad_norm": 6.731698550949781,
      "learning_rate": 5e-06,
      "loss": 0.1793,
      "step": 1080
    },
    {
      "epoch": 1.8734835355285961,
      "grad_norm": 12.021094536987366,
      "learning_rate": 5e-06,
      "loss": 0.336,
      "step": 1081
    },
    {
      "epoch": 1.8752166377816293,
      "grad_norm": 5.355346055020026,
      "learning_rate": 5e-06,
      "loss": 0.2119,
      "step": 1082
    },
    {
      "epoch": 1.8769497400346622,
      "grad_norm": 6.767457779864318,
      "learning_rate": 5e-06,
      "loss": 0.2305,
      "step": 1083
    },
    {
      "epoch": 1.878682842287695,
      "grad_norm": 4.787190057268374,
      "learning_rate": 5e-06,
      "loss": 0.1276,
      "step": 1084
    },
    {
      "epoch": 1.880415944540728,
      "grad_norm": 8.899824213585054,
      "learning_rate": 5e-06,
      "loss": 0.3599,
      "step": 1085
    },
    {
      "epoch": 1.882149046793761,
      "grad_norm": 8.281937250374494,
      "learning_rate": 5e-06,
      "loss": 0.3199,
      "step": 1086
    },
    {
      "epoch": 1.8838821490467939,
      "grad_norm": 4.652198411287227,
      "learning_rate": 5e-06,
      "loss": 0.1487,
      "step": 1087
    },
    {
      "epoch": 1.8856152512998268,
      "grad_norm": 9.880849144156098,
      "learning_rate": 5e-06,
      "loss": 0.3922,
      "step": 1088
    },
    {
      "epoch": 1.8873483535528597,
      "grad_norm": 7.470241509431172,
      "learning_rate": 5e-06,
      "loss": 0.1683,
      "step": 1089
    },
    {
      "epoch": 1.8890814558058926,
      "grad_norm": 7.341735823352836,
      "learning_rate": 5e-06,
      "loss": 0.1975,
      "step": 1090
    },
    {
      "epoch": 1.8908145580589255,
      "grad_norm": 8.847432478544402,
      "learning_rate": 5e-06,
      "loss": 0.2472,
      "step": 1091
    },
    {
      "epoch": 1.8925476603119584,
      "grad_norm": 12.210178344323857,
      "learning_rate": 5e-06,
      "loss": 0.324,
      "step": 1092
    },
    {
      "epoch": 1.8942807625649913,
      "grad_norm": 10.494599542996916,
      "learning_rate": 5e-06,
      "loss": 0.3004,
      "step": 1093
    },
    {
      "epoch": 1.8960138648180243,
      "grad_norm": 9.16415116445279,
      "learning_rate": 5e-06,
      "loss": 0.243,
      "step": 1094
    },
    {
      "epoch": 1.8977469670710572,
      "grad_norm": 12.771425085167852,
      "learning_rate": 5e-06,
      "loss": 0.4046,
      "step": 1095
    },
    {
      "epoch": 1.89948006932409,
      "grad_norm": 11.969915188876914,
      "learning_rate": 5e-06,
      "loss": 0.2209,
      "step": 1096
    },
    {
      "epoch": 1.901213171577123,
      "grad_norm": 7.293938186689015,
      "learning_rate": 5e-06,
      "loss": 0.226,
      "step": 1097
    },
    {
      "epoch": 1.902946273830156,
      "grad_norm": 14.107047496287095,
      "learning_rate": 5e-06,
      "loss": 0.305,
      "step": 1098
    },
    {
      "epoch": 1.9046793760831888,
      "grad_norm": 8.558163623804841,
      "learning_rate": 5e-06,
      "loss": 0.2158,
      "step": 1099
    },
    {
      "epoch": 1.9064124783362217,
      "grad_norm": 12.638419393871633,
      "learning_rate": 5e-06,
      "loss": 0.4498,
      "step": 1100
    },
    {
      "epoch": 1.9081455805892547,
      "grad_norm": 6.847490516480546,
      "learning_rate": 5e-06,
      "loss": 0.2147,
      "step": 1101
    },
    {
      "epoch": 1.9098786828422876,
      "grad_norm": 6.474652090262565,
      "learning_rate": 5e-06,
      "loss": 0.2904,
      "step": 1102
    },
    {
      "epoch": 1.9116117850953205,
      "grad_norm": 8.134386480135003,
      "learning_rate": 5e-06,
      "loss": 0.2252,
      "step": 1103
    },
    {
      "epoch": 1.9133448873483534,
      "grad_norm": 9.677379416106472,
      "learning_rate": 5e-06,
      "loss": 0.4385,
      "step": 1104
    },
    {
      "epoch": 1.9150779896013865,
      "grad_norm": 8.772253022781726,
      "learning_rate": 5e-06,
      "loss": 0.2563,
      "step": 1105
    },
    {
      "epoch": 1.9168110918544194,
      "grad_norm": 8.86382783269487,
      "learning_rate": 5e-06,
      "loss": 0.355,
      "step": 1106
    },
    {
      "epoch": 1.9185441941074524,
      "grad_norm": 5.596867550366256,
      "learning_rate": 5e-06,
      "loss": 0.1721,
      "step": 1107
    },
    {
      "epoch": 1.9202772963604853,
      "grad_norm": 7.186715257347097,
      "learning_rate": 5e-06,
      "loss": 0.2354,
      "step": 1108
    },
    {
      "epoch": 1.9220103986135182,
      "grad_norm": 8.095319794528416,
      "learning_rate": 5e-06,
      "loss": 0.299,
      "step": 1109
    },
    {
      "epoch": 1.923743500866551,
      "grad_norm": 7.678356408438577,
      "learning_rate": 5e-06,
      "loss": 0.2601,
      "step": 1110
    },
    {
      "epoch": 1.925476603119584,
      "grad_norm": 8.536730225294292,
      "learning_rate": 5e-06,
      "loss": 0.3303,
      "step": 1111
    },
    {
      "epoch": 1.927209705372617,
      "grad_norm": 7.62051384702663,
      "learning_rate": 5e-06,
      "loss": 0.2396,
      "step": 1112
    },
    {
      "epoch": 1.92894280762565,
      "grad_norm": 8.167773061530784,
      "learning_rate": 5e-06,
      "loss": 0.2953,
      "step": 1113
    },
    {
      "epoch": 1.930675909878683,
      "grad_norm": 7.027210482425202,
      "learning_rate": 5e-06,
      "loss": 0.3671,
      "step": 1114
    },
    {
      "epoch": 1.932409012131716,
      "grad_norm": 4.799038774884174,
      "learning_rate": 5e-06,
      "loss": 0.1541,
      "step": 1115
    },
    {
      "epoch": 1.9341421143847488,
      "grad_norm": 7.257359682540723,
      "learning_rate": 5e-06,
      "loss": 0.3218,
      "step": 1116
    },
    {
      "epoch": 1.9358752166377817,
      "grad_norm": 10.02793948488422,
      "learning_rate": 5e-06,
      "loss": 0.3249,
      "step": 1117
    },
    {
      "epoch": 1.9376083188908146,
      "grad_norm": 7.166497635141742,
      "learning_rate": 5e-06,
      "loss": 0.2128,
      "step": 1118
    },
    {
      "epoch": 1.9393414211438476,
      "grad_norm": 7.908562242942031,
      "learning_rate": 5e-06,
      "loss": 0.3676,
      "step": 1119
    },
    {
      "epoch": 1.9410745233968805,
      "grad_norm": 9.179979632984582,
      "learning_rate": 5e-06,
      "loss": 0.391,
      "step": 1120
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 2885,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 560,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.7899060793337446e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}