{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.11766090128250382,
  "eval_steps": 1000000,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00011766090128250382,
      "grad_norm": 0.0,
      "learning_rate": 0,
      "loss": 1.5093,
      "step": 1
    },
    {
      "epoch": 0.00023532180256500765,
      "grad_norm": 0.0,
      "learning_rate": 0,
      "loss": 1.6654,
      "step": 2
    },
    {
      "epoch": 0.00035298270384751147,
      "grad_norm": 7.219926357269287,
      "learning_rate": 0.0,
      "loss": 1.5056,
      "step": 3
    },
    {
      "epoch": 0.0004706436051300153,
      "grad_norm": 7.219926357269287,
      "learning_rate": 0.0,
      "loss": 1.6708,
      "step": 4
    },
    {
      "epoch": 0.0005883045064125191,
      "grad_norm": 5.301504135131836,
      "learning_rate": 1.5051499783199055e-07,
      "loss": 1.5035,
      "step": 5
    },
    {
      "epoch": 0.0007059654076950229,
      "grad_norm": 6.189344882965088,
      "learning_rate": 2.385606273598312e-07,
      "loss": 1.5641,
      "step": 6
    },
    {
      "epoch": 0.0008236263089775268,
      "grad_norm": 7.7528533935546875,
      "learning_rate": 3.010299956639811e-07,
      "loss": 1.6783,
      "step": 7
    },
    {
      "epoch": 0.0009412872102600306,
      "grad_norm": 5.569939613342285,
      "learning_rate": 3.494850021680093e-07,
      "loss": 1.4829,
      "step": 8
    },
    {
      "epoch": 0.0010589481115425344,
      "grad_norm": 5.5853753089904785,
      "learning_rate": 3.8907562519182173e-07,
      "loss": 1.505,
      "step": 9
    },
    {
      "epoch": 0.0011766090128250382,
      "grad_norm": 15.958133697509766,
      "learning_rate": 4.2254902000712834e-07,
      "loss": 1.5802,
      "step": 10
    },
    {
      "epoch": 0.001294269914107542,
      "grad_norm": 4.650132179260254,
      "learning_rate": 4.5154499349597166e-07,
      "loss": 1.4608,
      "step": 11
    },
    {
      "epoch": 0.0014119308153900459,
      "grad_norm": 8.008149147033691,
      "learning_rate": 4.771212547196623e-07,
      "loss": 1.5083,
      "step": 12
    },
    {
      "epoch": 0.0015295917166725497,
      "grad_norm": 9.279545783996582,
      "learning_rate": 4.999999999999999e-07,
      "loss": 1.4809,
      "step": 13
    },
    {
      "epoch": 0.0016472526179550535,
      "grad_norm": 3.9695677757263184,
      "learning_rate": 5.206963425791124e-07,
      "loss": 1.5041,
      "step": 14
    },
    {
      "epoch": 0.0017649135192375574,
      "grad_norm": 8.24997615814209,
      "learning_rate": 5.395906230238123e-07,
      "loss": 1.4886,
      "step": 15
    },
    {
      "epoch": 0.0018825744205200612,
      "grad_norm": 8.355944633483887,
      "learning_rate": 5.569716761534182e-07,
      "loss": 1.4472,
      "step": 16
    },
    {
      "epoch": 0.002000235321802565,
      "grad_norm": 6.1090474128723145,
      "learning_rate": 5.730640178391189e-07,
      "loss": 1.5176,
      "step": 17
    },
    {
      "epoch": 0.002117896223085069,
      "grad_norm": 6.527396202087402,
      "learning_rate": 5.880456295278405e-07,
      "loss": 1.6383,
      "step": 18
    },
    {
      "epoch": 0.0022355571243675727,
      "grad_norm": 5.9756693840026855,
      "learning_rate": 6.020599913279622e-07,
      "loss": 1.5845,
      "step": 19
    },
    {
      "epoch": 0.0023532180256500765,
      "grad_norm": 4.799884796142578,
      "learning_rate": 6.15224460689137e-07,
      "loss": 1.436,
      "step": 20
    },
    {
      "epoch": 0.0024708789269325803,
      "grad_norm": 7.455969333648682,
      "learning_rate": 6.276362525516529e-07,
      "loss": 1.6733,
      "step": 21
    },
    {
      "epoch": 0.002588539828215084,
      "grad_norm": 5.23336124420166,
      "learning_rate": 6.393768004764143e-07,
      "loss": 1.4561,
      "step": 22
    },
    {
      "epoch": 0.002706200729497588,
      "grad_norm": 6.086044788360596,
      "learning_rate": 6.505149978319905e-07,
      "loss": 1.5478,
      "step": 23
    },
    {
      "epoch": 0.0028238616307800918,
      "grad_norm": 7.104644298553467,
      "learning_rate": 6.611096473669595e-07,
      "loss": 1.5191,
      "step": 24
    },
    {
      "epoch": 0.0029415225320625956,
      "grad_norm": 5.2691521644592285,
      "learning_rate": 6.712113404111031e-07,
      "loss": 1.4956,
      "step": 25
    },
    {
      "epoch": 0.0030591834333450994,
      "grad_norm": 10.598758697509766,
      "learning_rate": 6.808639180087963e-07,
      "loss": 1.5545,
      "step": 26
    },
    {
      "epoch": 0.0031768443346276033,
      "grad_norm": 6.0853590965271,
      "learning_rate": 6.901056208558029e-07,
      "loss": 1.4893,
      "step": 27
    },
    {
      "epoch": 0.003294505235910107,
      "grad_norm": 4.166804313659668,
      "learning_rate": 6.989700043360186e-07,
      "loss": 1.4774,
      "step": 28
    },
    {
      "epoch": 0.003412166137192611,
      "grad_norm": 4.60942268371582,
      "learning_rate": 7.074866739854088e-07,
      "loss": 1.5448,
      "step": 29
    },
    {
      "epoch": 0.0035298270384751147,
      "grad_norm": 4.655628204345703,
      "learning_rate": 7.156818820794935e-07,
      "loss": 1.5081,
      "step": 30
    },
    {
      "epoch": 0.0036474879397576185,
      "grad_norm": 7.2692461013793945,
      "learning_rate": 7.235790156711094e-07,
      "loss": 1.5995,
      "step": 31
    },
    {
      "epoch": 0.0037651488410401224,
      "grad_norm": 4.613867282867432,
      "learning_rate": 7.311989989494779e-07,
      "loss": 1.4209,
      "step": 32
    },
    {
      "epoch": 0.003882809742322626,
      "grad_norm": 5.530818939208984,
      "learning_rate": 7.38560627359831e-07,
      "loss": 1.5738,
      "step": 33
    },
    {
      "epoch": 0.00400047064360513,
      "grad_norm": 4.4906487464904785,
      "learning_rate": 7.456808469171361e-07,
      "loss": 1.5313,
      "step": 34
    },
    {
      "epoch": 0.004118131544887634,
      "grad_norm": 6.488246440887451,
      "learning_rate": 7.525749891599529e-07,
      "loss": 1.405,
      "step": 35
    },
    {
      "epoch": 0.004235792446170138,
      "grad_norm": 6.306236267089844,
      "learning_rate": 7.592569699389436e-07,
      "loss": 1.5628,
      "step": 36
    },
    {
      "epoch": 0.0043534533474526415,
      "grad_norm": 6.571730136871338,
      "learning_rate": 7.657394585211274e-07,
      "loss": 1.5219,
      "step": 37
    },
    {
      "epoch": 0.004471114248735145,
      "grad_norm": 4.256100654602051,
      "learning_rate": 7.720340221751376e-07,
      "loss": 1.4473,
      "step": 38
    },
    {
      "epoch": 0.004588775150017649,
      "grad_norm": 9.501107215881348,
      "learning_rate": 7.781512503836435e-07,
      "loss": 1.4724,
      "step": 39
    },
    {
      "epoch": 0.004706436051300153,
      "grad_norm": 5.931169033050537,
      "learning_rate": 7.841008620334974e-07,
      "loss": 1.4485,
      "step": 40
    },
    {
      "epoch": 0.004824096952582657,
      "grad_norm": 4.35921049118042,
      "learning_rate": 7.89891798308405e-07,
      "loss": 1.4223,
      "step": 41
    },
    {
      "epoch": 0.004941757853865161,
      "grad_norm": 5.662957191467285,
      "learning_rate": 7.955323035132494e-07,
      "loss": 1.4246,
      "step": 42
    },
    {
      "epoch": 0.0050594187551476644,
      "grad_norm": 7.007707118988037,
      "learning_rate": 8.01029995663981e-07,
      "loss": 1.5855,
      "step": 43
    },
    {
      "epoch": 0.005177079656430168,
      "grad_norm": 4.280853748321533,
      "learning_rate": 8.063919283598676e-07,
      "loss": 1.3726,
      "step": 44
    },
    {
      "epoch": 0.005294740557712672,
      "grad_norm": 3.7020082473754883,
      "learning_rate": 8.116246451989502e-07,
      "loss": 1.3764,
      "step": 45
    },
    {
      "epoch": 0.005412401458995176,
      "grad_norm": 3.8952407836914062,
      "learning_rate": 8.16734227789793e-07,
      "loss": 1.3208,
      "step": 46
    },
    {
      "epoch": 0.00553006236027768,
      "grad_norm": 4.62631368637085,
      "learning_rate": 8.217263382430935e-07,
      "loss": 1.4958,
      "step": 47
    },
    {
      "epoch": 0.0056477232615601836,
      "grad_norm": 5.86137056350708,
      "learning_rate": 8.266062568876716e-07,
      "loss": 1.4349,
      "step": 48
    },
    {
      "epoch": 0.005765384162842687,
      "grad_norm": 4.414599895477295,
      "learning_rate": 8.313789158407869e-07,
      "loss": 1.3405,
      "step": 49
    },
    {
      "epoch": 0.005883045064125191,
      "grad_norm": 4.100174427032471,
      "learning_rate": 8.360489289678585e-07,
      "loss": 1.4465,
      "step": 50
    },
    {
      "epoch": 0.006000705965407695,
      "grad_norm": 4.100174427032471,
      "learning_rate": 8.360489289678585e-07,
      "loss": 1.5331,
      "step": 51
    },
    {
      "epoch": 0.006118366866690199,
      "grad_norm": 4.92950439453125,
      "learning_rate": 8.406206186877934e-07,
      "loss": 1.4608,
      "step": 52
    },
    {
      "epoch": 0.006236027767972703,
      "grad_norm": 3.7287988662719727,
      "learning_rate": 8.450980400142567e-07,
      "loss": 1.3522,
      "step": 53
    },
    {
      "epoch": 0.0063536886692552065,
      "grad_norm": 4.831335067749023,
      "learning_rate": 8.494850021680092e-07,
      "loss": 1.2695,
      "step": 54
    },
    {
      "epoch": 0.00647134957053771,
      "grad_norm": 4.615564346313477,
      "learning_rate": 8.53785088048968e-07,
      "loss": 1.4338,
      "step": 55
    },
    {
      "epoch": 0.006589010471820214,
      "grad_norm": 3.356062412261963,
      "learning_rate": 8.580016718173995e-07,
      "loss": 1.2757,
      "step": 56
    },
    {
      "epoch": 0.006706671373102718,
      "grad_norm": 5.449446678161621,
      "learning_rate": 8.621379348003944e-07,
      "loss": 1.356,
      "step": 57
    },
    {
      "epoch": 0.006824332274385222,
      "grad_norm": 3.9715359210968018,
      "learning_rate": 8.661968799114842e-07,
      "loss": 1.3804,
      "step": 58
    },
    {
      "epoch": 0.006941993175667726,
      "grad_norm": 3.3318145275115967,
      "learning_rate": 8.701813447471218e-07,
      "loss": 1.3677,
      "step": 59
    },
    {
      "epoch": 0.0070596540769502295,
      "grad_norm": 3.5082614421844482,
      "learning_rate": 8.740940135031001e-07,
      "loss": 1.2629,
      "step": 60
    },
    {
      "epoch": 0.007177314978232733,
      "grad_norm": 6.502823352813721,
      "learning_rate": 8.779374278362456e-07,
      "loss": 1.4748,
      "step": 61
    },
    {
      "epoch": 0.007294975879515237,
      "grad_norm": 4.117458820343018,
      "learning_rate": 8.817139967814684e-07,
      "loss": 1.2848,
      "step": 62
    },
    {
      "epoch": 0.007412636780797741,
      "grad_norm": 3.3300046920776367,
      "learning_rate": 8.854260058210719e-07,
      "loss": 1.3502,
      "step": 63
    },
    {
      "epoch": 0.007530297682080245,
      "grad_norm": 3.441394567489624,
      "learning_rate": 8.890756251918216e-07,
      "loss": 1.4513,
      "step": 64
    },
    {
      "epoch": 0.007647958583362749,
      "grad_norm": 4.066061019897461,
      "learning_rate": 8.926649175053833e-07,
      "loss": 1.2698,
      "step": 65
    },
    {
      "epoch": 0.007765619484645252,
      "grad_norm": 3.6549081802368164,
      "learning_rate": 8.961958447491268e-07,
      "loss": 1.2616,
      "step": 66
    },
    {
      "epoch": 0.007883280385927756,
      "grad_norm": 3.476713180541992,
      "learning_rate": 8.996702747267907e-07,
      "loss": 1.2418,
      "step": 67
    },
    {
      "epoch": 0.00800094128721026,
      "grad_norm": 4.902803421020508,
      "learning_rate": 9.030899869919433e-07,
      "loss": 1.448,
      "step": 68
    },
    {
      "epoch": 0.008118602188492764,
      "grad_norm": 4.095132827758789,
      "learning_rate": 9.064566783214276e-07,
      "loss": 1.3068,
      "step": 69
    },
    {
      "epoch": 0.008236263089775268,
      "grad_norm": 7.270352840423584,
      "learning_rate": 9.097719677709341e-07,
      "loss": 1.3759,
      "step": 70
    },
    {
      "epoch": 0.008353923991057772,
      "grad_norm": 11.875506401062012,
      "learning_rate": 9.13037401350413e-07,
      "loss": 1.3746,
      "step": 71
    },
    {
      "epoch": 0.008471584892340275,
      "grad_norm": 4.7003302574157715,
      "learning_rate": 9.162544563531181e-07,
      "loss": 1.2541,
      "step": 72
    },
    {
      "epoch": 0.00858924579362278,
      "grad_norm": 4.896581649780273,
      "learning_rate": 9.194245453686276e-07,
      "loss": 1.3253,
      "step": 73
    },
    {
      "epoch": 0.008706906694905283,
      "grad_norm": 4.074301242828369,
      "learning_rate": 9.225490200071283e-07,
      "loss": 1.2807,
      "step": 74
    },
    {
      "epoch": 0.008824567596187787,
      "grad_norm": 3.7367427349090576,
      "learning_rate": 9.256291743595375e-07,
      "loss": 1.3511,
      "step": 75
    },
    {
      "epoch": 0.00894222849747029,
      "grad_norm": 3.3777713775634766,
      "learning_rate": 9.28666248215634e-07,
      "loss": 1.2507,
      "step": 76
    },
    {
      "epoch": 0.009059889398752794,
      "grad_norm": 4.367025852203369,
      "learning_rate": 9.316614300602277e-07,
      "loss": 1.2705,
      "step": 77
    },
    {
      "epoch": 0.009177550300035298,
      "grad_norm": 5.1601362228393555,
      "learning_rate": 9.346158598654879e-07,
      "loss": 1.237,
      "step": 78
    },
    {
      "epoch": 0.009295211201317802,
      "grad_norm": 6.360751152038574,
      "learning_rate": 9.375306316958498e-07,
      "loss": 1.2599,
      "step": 79
    },
    {
      "epoch": 0.009412872102600306,
      "grad_norm": 3.5460028648376465,
      "learning_rate": 9.404067961403955e-07,
      "loss": 1.1984,
      "step": 80
    },
    {
      "epoch": 0.00953053300388281,
      "grad_norm": 3.6358044147491455,
      "learning_rate": 9.432453625862408e-07,
      "loss": 1.3739,
      "step": 81
    },
    {
      "epoch": 0.009648193905165314,
      "grad_norm": 3.501857042312622,
      "learning_rate": 9.4604730134524e-07,
      "loss": 1.2913,
      "step": 82
    },
    {
      "epoch": 0.009765854806447817,
      "grad_norm": 4.328304767608643,
      "learning_rate": 9.488135456452205e-07,
      "loss": 1.2586,
      "step": 83
    },
    {
      "epoch": 0.009883515707730321,
      "grad_norm": 4.727443218231201,
      "learning_rate": 9.515449934959715e-07,
      "loss": 1.2806,
      "step": 84
    },
    {
      "epoch": 0.010001176609012825,
      "grad_norm": 2.271977186203003,
      "learning_rate": 9.542425094393247e-07,
      "loss": 1.2396,
      "step": 85
    },
    {
      "epoch": 0.010118837510295329,
      "grad_norm": 5.965804100036621,
      "learning_rate": 9.569069261918583e-07,
      "loss": 1.2298,
      "step": 86
    },
    {
      "epoch": 0.010236498411577833,
      "grad_norm": 3.108452081680298,
      "learning_rate": 9.59539046188037e-07,
      "loss": 1.3033,
      "step": 87
    },
    {
      "epoch": 0.010354159312860337,
      "grad_norm": 3.3207242488861084,
      "learning_rate": 9.621396430309406e-07,
      "loss": 1.2343,
      "step": 88
    },
    {
      "epoch": 0.01047182021414284,
      "grad_norm": 4.672191619873047,
      "learning_rate": 9.647094628571462e-07,
      "loss": 1.2168,
      "step": 89
    },
    {
      "epoch": 0.010589481115425344,
      "grad_norm": 5.991580963134766,
      "learning_rate": 9.672492256217836e-07,
      "loss": 1.2139,
      "step": 90
    },
    {
      "epoch": 0.010707142016707848,
      "grad_norm": 3.4057729244232178,
      "learning_rate": 9.69759626309309e-07,
      "loss": 1.3043,
      "step": 91
    },
    {
      "epoch": 0.010824802917990352,
      "grad_norm": 2.9600138664245605,
      "learning_rate": 9.722413360750842e-07,
      "loss": 1.2457,
      "step": 92
    },
    {
      "epoch": 0.010942463819272856,
      "grad_norm": 5.642834663391113,
      "learning_rate": 9.74695003322456e-07,
      "loss": 1.2287,
      "step": 93
    },
    {
      "epoch": 0.01106012472055536,
      "grad_norm": 3.348522186279297,
      "learning_rate": 9.771212547196622e-07,
      "loss": 1.2436,
      "step": 94
    },
    {
      "epoch": 0.011177785621837863,
      "grad_norm": 2.7108304500579834,
      "learning_rate": 9.795206961605466e-07,
      "loss": 1.2419,
      "step": 95
    },
    {
      "epoch": 0.011295446523120367,
      "grad_norm": 4.255400657653809,
      "learning_rate": 9.818939136727774e-07,
      "loss": 1.2603,
      "step": 96
    },
    {
      "epoch": 0.011413107424402871,
      "grad_norm": 3.6323165893554688,
      "learning_rate": 9.842414742769674e-07,
      "loss": 1.2927,
      "step": 97
    },
    {
      "epoch": 0.011530768325685375,
      "grad_norm": 8.131080627441406,
      "learning_rate": 9.865639267998492e-07,
      "loss": 1.1739,
      "step": 98
    },
    {
      "epoch": 0.011648429226967879,
      "grad_norm": 3.69378662109375,
      "learning_rate": 9.888618026444236e-07,
      "loss": 1.2147,
      "step": 99
    },
    {
      "epoch": 0.011766090128250382,
      "grad_norm": 3.5188376903533936,
      "learning_rate": 9.91135616519784e-07,
      "loss": 1.1089,
      "step": 100
    },
    {
      "epoch": 0.011883751029532886,
      "grad_norm": 3.0614805221557617,
      "learning_rate": 9.933858671331222e-07,
      "loss": 1.2198,
      "step": 101
    },
    {
      "epoch": 0.01200141193081539,
      "grad_norm": 3.6954472064971924,
      "learning_rate": 9.956130378462473e-07,
      "loss": 1.1984,
      "step": 102
    },
    {
      "epoch": 0.012119072832097894,
      "grad_norm": 2.885023593902588,
      "learning_rate": 9.978175972987748e-07,
      "loss": 1.2223,
      "step": 103
    },
    {
      "epoch": 0.012236733733380398,
      "grad_norm": 5.114850044250488,
      "learning_rate": 9.999999999999997e-07,
      "loss": 1.3275,
      "step": 104
    },
    {
      "epoch": 0.012354394634662902,
      "grad_norm": 2.7323529720306396,
      "learning_rate": 1e-06,
      "loss": 1.1721,
      "step": 105
    },
    {
      "epoch": 0.012472055535945405,
      "grad_norm": 3.372161626815796,
      "learning_rate": 1e-06,
      "loss": 1.1658,
      "step": 106
    },
    {
      "epoch": 0.01258971643722791,
      "grad_norm": 3.060661554336548,
      "learning_rate": 1e-06,
      "loss": 1.181,
      "step": 107
    },
    {
      "epoch": 0.012707377338510413,
      "grad_norm": 4.042975425720215,
      "learning_rate": 1e-06,
      "loss": 1.2399,
      "step": 108
    },
    {
      "epoch": 0.012825038239792917,
      "grad_norm": 2.5195248126983643,
      "learning_rate": 1e-06,
      "loss": 1.1678,
      "step": 109
    },
    {
      "epoch": 0.01294269914107542,
      "grad_norm": 2.3371739387512207,
      "learning_rate": 1e-06,
      "loss": 1.0537,
      "step": 110
    },
    {
      "epoch": 0.013060360042357924,
      "grad_norm": 3.8651785850524902,
      "learning_rate": 1e-06,
      "loss": 1.227,
      "step": 111
    },
    {
      "epoch": 0.013178020943640428,
      "grad_norm": 3.3304595947265625,
      "learning_rate": 1e-06,
      "loss": 1.1826,
      "step": 112
    },
    {
      "epoch": 0.013295681844922932,
      "grad_norm": 5.385232448577881,
      "learning_rate": 1e-06,
      "loss": 1.2031,
      "step": 113
    },
    {
      "epoch": 0.013413342746205436,
      "grad_norm": 3.4247732162475586,
      "learning_rate": 1e-06,
      "loss": 1.2171,
      "step": 114
    },
    {
      "epoch": 0.01353100364748794,
      "grad_norm": 4.058660507202148,
      "learning_rate": 1e-06,
      "loss": 1.2008,
      "step": 115
    },
    {
      "epoch": 0.013648664548770444,
      "grad_norm": 7.034421443939209,
      "learning_rate": 1e-06,
      "loss": 1.1456,
      "step": 116
    },
    {
      "epoch": 0.013766325450052947,
      "grad_norm": 6.617616653442383,
      "learning_rate": 1e-06,
      "loss": 1.26,
      "step": 117
    },
    {
      "epoch": 0.013883986351335451,
      "grad_norm": 1.7746769189834595,
      "learning_rate": 1e-06,
      "loss": 1.1115,
      "step": 118
    },
    {
      "epoch": 0.014001647252617955,
      "grad_norm": 5.424281120300293,
      "learning_rate": 1e-06,
      "loss": 1.231,
      "step": 119
    },
    {
      "epoch": 0.014119308153900459,
      "grad_norm": 4.011491775512695,
      "learning_rate": 1e-06,
      "loss": 1.1399,
      "step": 120
    },
    {
      "epoch": 0.014236969055182963,
      "grad_norm": 3.085541009902954,
      "learning_rate": 1e-06,
      "loss": 1.1943,
      "step": 121
    },
    {
      "epoch": 0.014354629956465467,
      "grad_norm": 5.818262577056885,
      "learning_rate": 1e-06,
      "loss": 1.3063,
      "step": 122
    },
    {
      "epoch": 0.01447229085774797,
      "grad_norm": 2.9598960876464844,
      "learning_rate": 1e-06,
      "loss": 1.143,
      "step": 123
    },
    {
      "epoch": 0.014589951759030474,
      "grad_norm": 2.9467811584472656,
      "learning_rate": 1e-06,
      "loss": 1.2684,
      "step": 124
    },
    {
      "epoch": 0.014707612660312978,
      "grad_norm": 2.971536159515381,
      "learning_rate": 1e-06,
      "loss": 1.2013,
      "step": 125
    },
    {
      "epoch": 0.014825273561595482,
      "grad_norm": 3.8481204509735107,
      "learning_rate": 1e-06,
      "loss": 1.1681,
      "step": 126
    },
    {
      "epoch": 0.014942934462877986,
      "grad_norm": 2.742084503173828,
      "learning_rate": 1e-06,
      "loss": 1.1682,
      "step": 127
    },
    {
      "epoch": 0.01506059536416049,
      "grad_norm": 9.748570442199707,
      "learning_rate": 1e-06,
      "loss": 1.1744,
      "step": 128
    },
    {
      "epoch": 0.015178256265442993,
      "grad_norm": 2.9641549587249756,
      "learning_rate": 1e-06,
      "loss": 1.1502,
      "step": 129
    },
    {
      "epoch": 0.015295917166725497,
      "grad_norm": 2.597259044647217,
      "learning_rate": 1e-06,
      "loss": 1.0995,
      "step": 130
    },
    {
      "epoch": 0.015413578068008001,
      "grad_norm": 4.812168121337891,
      "learning_rate": 1e-06,
      "loss": 1.1822,
      "step": 131
    },
    {
      "epoch": 0.015531238969290505,
      "grad_norm": 3.0024499893188477,
      "learning_rate": 1e-06,
      "loss": 1.2001,
      "step": 132
    },
    {
      "epoch": 0.015648899870573007,
      "grad_norm": 1.608038306236267,
      "learning_rate": 1e-06,
      "loss": 1.1332,
      "step": 133
    },
    {
      "epoch": 0.015766560771855512,
      "grad_norm": 1.6776491403579712,
      "learning_rate": 1e-06,
      "loss": 1.0944,
      "step": 134
    },
    {
      "epoch": 0.015884221673138015,
      "grad_norm": 2.8363330364227295,
      "learning_rate": 1e-06,
      "loss": 1.1976,
      "step": 135
    },
    {
      "epoch": 0.01600188257442052,
      "grad_norm": 3.921804904937744,
      "learning_rate": 1e-06,
      "loss": 1.1713,
      "step": 136
    },
    {
      "epoch": 0.016119543475703022,
      "grad_norm": 4.1015777587890625,
      "learning_rate": 1e-06,
      "loss": 1.1039,
      "step": 137
    },
    {
      "epoch": 0.016237204376985528,
      "grad_norm": 2.807033061981201,
      "learning_rate": 1e-06,
      "loss": 1.1538,
      "step": 138
    },
    {
      "epoch": 0.01635486527826803,
      "grad_norm": 3.5602121353149414,
      "learning_rate": 1e-06,
      "loss": 1.1008,
      "step": 139
    },
    {
      "epoch": 0.016472526179550535,
      "grad_norm": 2.0294995307922363,
      "learning_rate": 1e-06,
      "loss": 1.1416,
      "step": 140
    },
    {
      "epoch": 0.016590187080833037,
      "grad_norm": 2.332132339477539,
      "learning_rate": 1e-06,
      "loss": 1.0754,
      "step": 141
    },
    {
      "epoch": 0.016707847982115543,
      "grad_norm": 3.4389145374298096,
      "learning_rate": 1e-06,
      "loss": 1.1983,
      "step": 142
    },
    {
      "epoch": 0.016825508883398045,
      "grad_norm": 3.4869580268859863,
      "learning_rate": 1e-06,
      "loss": 1.0771,
      "step": 143
    },
    {
      "epoch": 0.01694316978468055,
      "grad_norm": 3.293520212173462,
      "learning_rate": 1e-06,
      "loss": 1.0733,
      "step": 144
    },
    {
      "epoch": 0.017060830685963053,
      "grad_norm": 3.1414260864257812,
      "learning_rate": 1e-06,
      "loss": 1.2403,
      "step": 145
    },
    {
      "epoch": 0.01717849158724556,
      "grad_norm": 2.8440492153167725,
      "learning_rate": 1e-06,
      "loss": 1.153,
      "step": 146
    },
    {
      "epoch": 0.01729615248852806,
      "grad_norm": 3.1873555183410645,
      "learning_rate": 1e-06,
      "loss": 1.192,
      "step": 147
    },
    {
      "epoch": 0.017413813389810566,
      "grad_norm": 4.0402069091796875,
      "learning_rate": 1e-06,
      "loss": 1.0346,
      "step": 148
    },
    {
      "epoch": 0.017531474291093068,
      "grad_norm": 2.9087209701538086,
      "learning_rate": 1e-06,
      "loss": 1.1442,
      "step": 149
    },
    {
      "epoch": 0.017649135192375574,
      "grad_norm": 2.827258825302124,
      "learning_rate": 1e-06,
      "loss": 1.0892,
      "step": 150
    },
    {
      "epoch": 0.017766796093658076,
      "grad_norm": 4.406788349151611,
      "learning_rate": 1e-06,
      "loss": 1.135,
      "step": 151
    },
    {
      "epoch": 0.01788445699494058,
      "grad_norm": 5.041503429412842,
      "learning_rate": 1e-06,
      "loss": 1.1424,
      "step": 152
    },
    {
      "epoch": 0.018002117896223083,
      "grad_norm": 2.4025824069976807,
      "learning_rate": 1e-06,
      "loss": 1.1524,
      "step": 153
    },
    {
      "epoch": 0.01811977879750559,
      "grad_norm": 4.14438009262085,
      "learning_rate": 1e-06,
      "loss": 1.1283,
      "step": 154
    },
    {
      "epoch": 0.01823743969878809,
      "grad_norm": 3.5507988929748535,
      "learning_rate": 1e-06,
      "loss": 1.0656,
      "step": 155
    },
    {
      "epoch": 0.018355100600070597,
      "grad_norm": 2.2941172122955322,
      "learning_rate": 1e-06,
      "loss": 1.0974,
      "step": 156
    },
    {
      "epoch": 0.0184727615013531,
      "grad_norm": 2.8862037658691406,
      "learning_rate": 1e-06,
      "loss": 1.1994,
      "step": 157
    },
    {
      "epoch": 0.018590422402635604,
      "grad_norm": 4.5398054122924805,
      "learning_rate": 1e-06,
      "loss": 1.0714,
      "step": 158
    },
    {
      "epoch": 0.018708083303918106,
      "grad_norm": 2.8482394218444824,
      "learning_rate": 1e-06,
      "loss": 1.1728,
      "step": 159
    },
    {
      "epoch": 0.018825744205200612,
      "grad_norm": 3.723759412765503,
      "learning_rate": 1e-06,
      "loss": 1.1599,
      "step": 160
    },
    {
      "epoch": 0.018943405106483114,
      "grad_norm": 2.3328919410705566,
      "learning_rate": 1e-06,
      "loss": 1.0474,
      "step": 161
    },
    {
      "epoch": 0.01906106600776562,
      "grad_norm": 3.5935370922088623,
      "learning_rate": 1e-06,
      "loss": 1.1742,
      "step": 162
    },
    {
      "epoch": 0.01917872690904812,
      "grad_norm": 4.418773651123047,
      "learning_rate": 1e-06,
      "loss": 1.1212,
      "step": 163
    },
    {
      "epoch": 0.019296387810330627,
      "grad_norm": 2.6029717922210693,
      "learning_rate": 1e-06,
      "loss": 1.0893,
      "step": 164
    },
    {
      "epoch": 0.01941404871161313,
      "grad_norm": 5.034754753112793,
      "learning_rate": 1e-06,
      "loss": 1.0841,
      "step": 165
    },
    {
      "epoch": 0.019531709612895635,
      "grad_norm": 3.910154104232788,
      "learning_rate": 1e-06,
      "loss": 1.0441,
      "step": 166
    },
    {
      "epoch": 0.019649370514178137,
      "grad_norm": 5.273410797119141,
      "learning_rate": 1e-06,
      "loss": 1.1746,
      "step": 167
    },
    {
      "epoch": 0.019767031415460642,
      "grad_norm": 12.92271614074707,
      "learning_rate": 1e-06,
      "loss": 1.082,
      "step": 168
    },
    {
      "epoch": 0.019884692316743145,
      "grad_norm": 4.1488356590271,
      "learning_rate": 1e-06,
      "loss": 1.1214,
      "step": 169
    },
    {
      "epoch": 0.02000235321802565,
      "grad_norm": 2.6337387561798096,
      "learning_rate": 1e-06,
      "loss": 1.1608,
      "step": 170
    },
    {
      "epoch": 0.020120014119308152,
      "grad_norm": 4.516927242279053,
      "learning_rate": 1e-06,
      "loss": 1.0716,
      "step": 171
    },
    {
      "epoch": 0.020237675020590658,
      "grad_norm": 3.008997678756714,
      "learning_rate": 1e-06,
      "loss": 1.0661,
      "step": 172
    },
    {
      "epoch": 0.02035533592187316,
      "grad_norm": 2.972275495529175,
      "learning_rate": 1e-06,
      "loss": 1.0881,
      "step": 173
    },
    {
      "epoch": 0.020472996823155665,
      "grad_norm": 3.5600616931915283,
      "learning_rate": 1e-06,
      "loss": 1.0664,
      "step": 174
    },
    {
      "epoch": 0.020590657724438167,
      "grad_norm": 4.367973327636719,
      "learning_rate": 1e-06,
      "loss": 1.0726,
      "step": 175
    },
    {
      "epoch": 0.020708318625720673,
      "grad_norm": 3.5119714736938477,
      "learning_rate": 1e-06,
      "loss": 1.1287,
      "step": 176
    },
    {
      "epoch": 0.020825979527003175,
      "grad_norm": 1.7865749597549438,
      "learning_rate": 1e-06,
      "loss": 1.0975,
      "step": 177
    },
    {
      "epoch": 0.02094364042828568,
      "grad_norm": 3.772221088409424,
      "learning_rate": 1e-06,
      "loss": 1.0591,
      "step": 178
    },
    {
      "epoch": 0.021061301329568183,
      "grad_norm": 3.0297372341156006,
      "learning_rate": 1e-06,
      "loss": 1.0511,
      "step": 179
    },
    {
      "epoch": 0.02117896223085069,
      "grad_norm": 3.609771728515625,
      "learning_rate": 1e-06,
      "loss": 1.0995,
      "step": 180
    },
    {
      "epoch": 0.02129662313213319,
      "grad_norm": 3.8480026721954346,
      "learning_rate": 1e-06,
      "loss": 1.0532,
      "step": 181
    },
    {
      "epoch": 0.021414284033415696,
      "grad_norm": 4.087072849273682,
      "learning_rate": 1e-06,
      "loss": 1.0413,
      "step": 182
    },
    {
      "epoch": 0.021531944934698198,
      "grad_norm": 1.8740599155426025,
      "learning_rate": 1e-06,
      "loss": 1.0721,
      "step": 183
    },
    {
      "epoch": 0.021649605835980704,
      "grad_norm": 4.0999436378479,
      "learning_rate": 1e-06,
      "loss": 1.1368,
      "step": 184
    },
    {
      "epoch": 0.021767266737263206,
      "grad_norm": 2.669431686401367,
      "learning_rate": 1e-06,
      "loss": 1.1078,
      "step": 185
    },
    {
      "epoch": 0.02188492763854571,
      "grad_norm": 2.4161550998687744,
      "learning_rate": 1e-06,
      "loss": 0.9679,
      "step": 186
    },
    {
      "epoch": 0.022002588539828213,
      "grad_norm": 3.029654026031494,
      "learning_rate": 1e-06,
      "loss": 1.0193,
      "step": 187
    },
    {
      "epoch": 0.02212024944111072,
      "grad_norm": 3.7600207328796387,
      "learning_rate": 1e-06,
      "loss": 1.1391,
      "step": 188
    },
    {
      "epoch": 0.02223791034239322,
      "grad_norm": 2.874558210372925,
      "learning_rate": 1e-06,
      "loss": 1.0098,
      "step": 189
    },
    {
      "epoch": 0.022355571243675727,
      "grad_norm": 6.042860507965088,
      "learning_rate": 1e-06,
      "loss": 0.9819,
      "step": 190
    },
    {
      "epoch": 0.02247323214495823,
      "grad_norm": 1.992511510848999,
      "learning_rate": 1e-06,
      "loss": 0.9834,
      "step": 191
    },
    {
      "epoch": 0.022590893046240734,
      "grad_norm": 6.9423065185546875,
      "learning_rate": 1e-06,
      "loss": 1.0761,
      "step": 192
    },
    {
      "epoch": 0.022708553947523236,
      "grad_norm": 1.9064056873321533,
      "learning_rate": 1e-06,
      "loss": 0.9993,
      "step": 193
    },
    {
      "epoch": 0.022826214848805742,
      "grad_norm": 2.7809596061706543,
      "learning_rate": 1e-06,
      "loss": 1.1212,
      "step": 194
    },
    {
      "epoch": 0.022943875750088244,
      "grad_norm": 3.407334804534912,
      "learning_rate": 1e-06,
      "loss": 1.0231,
      "step": 195
    },
    {
      "epoch": 0.02306153665137075,
      "grad_norm": 3.5821261405944824,
      "learning_rate": 1e-06,
      "loss": 0.9963,
      "step": 196
    },
    {
      "epoch": 0.02317919755265325,
      "grad_norm": 3.629689931869507,
      "learning_rate": 1e-06,
      "loss": 1.0986,
      "step": 197
    },
    {
      "epoch": 0.023296858453935757,
      "grad_norm": 9.692272186279297,
      "learning_rate": 1e-06,
      "loss": 1.0639,
      "step": 198
    },
    {
      "epoch": 0.02341451935521826,
      "grad_norm": 2.2132132053375244,
      "learning_rate": 1e-06,
      "loss": 0.967,
      "step": 199
    },
    {
      "epoch": 0.023532180256500765,
      "grad_norm": 3.079921007156372,
      "learning_rate": 1e-06,
      "loss": 1.004,
      "step": 200
    },
    {
      "epoch": 0.023649841157783267,
      "grad_norm": 3.2911813259124756,
      "learning_rate": 1e-06,
      "loss": 1.0534,
      "step": 201
    },
    {
      "epoch": 0.023767502059065772,
      "grad_norm": 4.560142517089844,
      "learning_rate": 1e-06,
      "loss": 0.9946,
      "step": 202
    },
    {
      "epoch": 0.023885162960348275,
      "grad_norm": 2.6420490741729736,
      "learning_rate": 1e-06,
      "loss": 1.0177,
      "step": 203
    },
    {
      "epoch": 0.02400282386163078,
      "grad_norm": 2.756321668624878,
      "learning_rate": 1e-06,
      "loss": 1.0444,
      "step": 204
    },
    {
      "epoch": 0.024120484762913282,
      "grad_norm": 2.4896962642669678,
      "learning_rate": 1e-06,
      "loss": 1.0205,
      "step": 205
    },
    {
      "epoch": 0.024238145664195788,
      "grad_norm": 10.722167015075684,
      "learning_rate": 1e-06,
      "loss": 1.0348,
      "step": 206
    },
    {
      "epoch": 0.02435580656547829,
      "grad_norm": 2.6661248207092285,
      "learning_rate": 1e-06,
      "loss": 1.0385,
      "step": 207
    },
    {
      "epoch": 0.024473467466760795,
      "grad_norm": 4.628503799438477,
      "learning_rate": 1e-06,
      "loss": 0.972,
      "step": 208
    },
    {
      "epoch": 0.024591128368043298,
      "grad_norm": 3.111830234527588,
      "learning_rate": 1e-06,
      "loss": 1.1072,
      "step": 209
    },
    {
      "epoch": 0.024708789269325803,
      "grad_norm": 4.0781402587890625,
      "learning_rate": 1e-06,
      "loss": 1.0554,
      "step": 210
    },
    {
      "epoch": 0.024826450170608305,
      "grad_norm": 5.678940773010254,
      "learning_rate": 1e-06,
      "loss": 1.1226,
      "step": 211
    },
    {
      "epoch": 0.02494411107189081,
      "grad_norm": 5.578497886657715,
      "learning_rate": 1e-06,
      "loss": 0.9829,
      "step": 212
    },
    {
      "epoch": 0.025061771973173313,
      "grad_norm": 2.8906311988830566,
      "learning_rate": 1e-06,
      "loss": 0.9563,
      "step": 213
    },
    {
      "epoch": 0.02517943287445582,
      "grad_norm": 3.579702138900757,
      "learning_rate": 1e-06,
      "loss": 1.0379,
      "step": 214
    },
    {
      "epoch": 0.02529709377573832,
      "grad_norm": 3.536996841430664,
      "learning_rate": 1e-06,
      "loss": 0.9938,
      "step": 215
    },
    {
      "epoch": 0.025414754677020826,
      "grad_norm": 2.8766660690307617,
      "learning_rate": 1e-06,
      "loss": 0.9176,
      "step": 216
    },
    {
      "epoch": 0.025532415578303328,
      "grad_norm": 6.301769256591797,
      "learning_rate": 1e-06,
      "loss": 1.1188,
      "step": 217
    },
    {
      "epoch": 0.025650076479585834,
      "grad_norm": 6.618703365325928,
      "learning_rate": 1e-06,
      "loss": 1.0338,
      "step": 218
    },
    {
      "epoch": 0.025767737380868336,
      "grad_norm": 3.9064433574676514,
      "learning_rate": 1e-06,
      "loss": 0.9414,
      "step": 219
    },
    {
      "epoch": 0.02588539828215084,
      "grad_norm": 2.456340789794922,
      "learning_rate": 1e-06,
      "loss": 1.0252,
      "step": 220
    },
    {
      "epoch": 0.026003059183433343,
      "grad_norm": 3.485226631164551,
      "learning_rate": 1e-06,
      "loss": 1.0534,
      "step": 221
    },
    {
      "epoch": 0.02612072008471585,
      "grad_norm": 4.325056552886963,
      "learning_rate": 1e-06,
      "loss": 1.0813,
      "step": 222
    },
    {
      "epoch": 0.02623838098599835,
      "grad_norm": 3.2998199462890625,
      "learning_rate": 1e-06,
      "loss": 1.0372,
      "step": 223
    },
    {
      "epoch": 0.026356041887280857,
      "grad_norm": 3.8687708377838135,
      "learning_rate": 1e-06,
      "loss": 0.9828,
      "step": 224
    },
    {
      "epoch": 0.02647370278856336,
      "grad_norm": 3.572066307067871,
      "learning_rate": 1e-06,
      "loss": 1.0074,
      "step": 225
    },
    {
      "epoch": 0.026591363689845864,
      "grad_norm": 2.7380471229553223,
      "learning_rate": 1e-06,
      "loss": 0.8806,
      "step": 226
    },
    {
      "epoch": 0.026709024591128366,
      "grad_norm": 4.645740032196045,
      "learning_rate": 1e-06,
      "loss": 1.0778,
      "step": 227
    },
    {
      "epoch": 0.026826685492410872,
      "grad_norm": 4.34233283996582,
      "learning_rate": 1e-06,
      "loss": 1.0747,
      "step": 228
    },
    {
      "epoch": 0.026944346393693374,
      "grad_norm": 3.2557373046875,
      "learning_rate": 1e-06,
      "loss": 0.9457,
      "step": 229
    },
    {
      "epoch": 0.02706200729497588,
      "grad_norm": 4.202934741973877,
      "learning_rate": 1e-06,
      "loss": 0.8844,
      "step": 230
    },
    {
      "epoch": 0.02717966819625838,
      "grad_norm": 2.7801413536071777,
      "learning_rate": 1e-06,
      "loss": 1.0413,
      "step": 231
    },
    {
      "epoch": 0.027297329097540887,
      "grad_norm": 3.5069901943206787,
      "learning_rate": 1e-06,
      "loss": 1.0001,
      "step": 232
    },
    {
      "epoch": 0.02741498999882339,
      "grad_norm": 3.3977630138397217,
      "learning_rate": 1e-06,
      "loss": 1.0629,
      "step": 233
    },
    {
      "epoch": 0.027532650900105895,
      "grad_norm": 2.3797481060028076,
      "learning_rate": 1e-06,
      "loss": 0.9498,
      "step": 234
    },
    {
      "epoch": 0.027650311801388397,
      "grad_norm": 2.1209325790405273,
      "learning_rate": 1e-06,
      "loss": 1.2058,
      "step": 235
    },
    {
      "epoch": 0.027767972702670903,
      "grad_norm": 3.580355644226074,
      "learning_rate": 1e-06,
      "loss": 1.0018,
      "step": 236
    },
    {
      "epoch": 0.027885633603953405,
      "grad_norm": 4.806403160095215,
      "learning_rate": 1e-06,
      "loss": 1.0458,
      "step": 237
    },
    {
      "epoch": 0.02800329450523591,
      "grad_norm": 2.399860382080078,
      "learning_rate": 1e-06,
      "loss": 0.9849,
      "step": 238
    },
    {
      "epoch": 0.028120955406518412,
      "grad_norm": 3.2041261196136475,
      "learning_rate": 1e-06,
      "loss": 1.0088,
      "step": 239
    },
    {
      "epoch": 0.028238616307800918,
      "grad_norm": 5.033557891845703,
      "learning_rate": 1e-06,
      "loss": 1.1207,
      "step": 240
    },
    {
      "epoch": 0.02835627720908342,
      "grad_norm": 2.3300740718841553,
      "learning_rate": 1e-06,
      "loss": 0.987,
      "step": 241
    },
    {
      "epoch": 0.028473938110365925,
      "grad_norm": 3.4651834964752197,
      "learning_rate": 1e-06,
      "loss": 0.9708,
      "step": 242
    },
    {
      "epoch": 0.028591599011648428,
      "grad_norm": 5.092766284942627,
      "learning_rate": 1e-06,
      "loss": 0.9562,
      "step": 243
    },
    {
      "epoch": 0.028709259912930933,
      "grad_norm": 3.308539390563965,
      "learning_rate": 1e-06,
      "loss": 1.0092,
      "step": 244
    },
    {
      "epoch": 0.028826920814213435,
      "grad_norm": 2.86190128326416,
      "learning_rate": 1e-06,
      "loss": 0.9801,
      "step": 245
    },
    {
      "epoch": 0.02894458171549594,
      "grad_norm": 2.9260149002075195,
      "learning_rate": 1e-06,
      "loss": 0.8868,
      "step": 246
    },
    {
      "epoch": 0.029062242616778443,
      "grad_norm": 5.265139102935791,
      "learning_rate": 1e-06,
      "loss": 0.9974,
      "step": 247
    },
    {
      "epoch": 0.02917990351806095,
      "grad_norm": 5.411443710327148,
      "learning_rate": 1e-06,
      "loss": 1.0284,
      "step": 248
    },
    {
      "epoch": 0.02929756441934345,
      "grad_norm": 4.155786037445068,
      "learning_rate": 1e-06,
      "loss": 1.0795,
      "step": 249
    },
    {
      "epoch": 0.029415225320625956,
      "grad_norm": 2.910003900527954,
      "learning_rate": 1e-06,
      "loss": 1.0173,
      "step": 250
    },
    {
      "epoch": 0.029532886221908458,
      "grad_norm": 2.214527130126953,
      "learning_rate": 1e-06,
      "loss": 0.98,
      "step": 251
    },
    {
      "epoch": 0.029650547123190964,
      "grad_norm": 6.6619086265563965,
      "learning_rate": 1e-06,
      "loss": 0.9513,
      "step": 252
    },
    {
      "epoch": 0.029768208024473466,
      "grad_norm": 2.5836620330810547,
      "learning_rate": 1e-06,
      "loss": 0.9591,
      "step": 253
    },
    {
      "epoch": 0.02988586892575597,
      "grad_norm": 3.5939230918884277,
      "learning_rate": 1e-06,
      "loss": 0.9877,
      "step": 254
    },
    {
      "epoch": 0.030003529827038473,
      "grad_norm": 3.4384496212005615,
      "learning_rate": 1e-06,
      "loss": 0.8965,
      "step": 255
    },
    {
      "epoch": 0.03012119072832098,
      "grad_norm": 5.588817119598389,
      "learning_rate": 1e-06,
      "loss": 0.9589,
      "step": 256
    },
    {
      "epoch": 0.03023885162960348,
      "grad_norm": 5.522799015045166,
      "learning_rate": 1e-06,
      "loss": 1.0202,
      "step": 257
    },
    {
      "epoch": 0.030356512530885987,
      "grad_norm": 3.071120500564575,
      "learning_rate": 1e-06,
      "loss": 0.9195,
      "step": 258
    },
    {
      "epoch": 0.03047417343216849,
      "grad_norm": 8.155257225036621,
      "learning_rate": 1e-06,
      "loss": 0.9719,
      "step": 259
    },
    {
      "epoch": 0.030591834333450994,
      "grad_norm": 2.8112828731536865,
      "learning_rate": 1e-06,
      "loss": 0.9657,
      "step": 260
    },
    {
      "epoch": 0.030709495234733496,
      "grad_norm": 6.376706600189209,
      "learning_rate": 1e-06,
      "loss": 0.9453,
      "step": 261
    },
    {
      "epoch": 0.030827156136016002,
      "grad_norm": 2.884045124053955,
      "learning_rate": 1e-06,
      "loss": 0.8969,
      "step": 262
    },
    {
      "epoch": 0.030944817037298504,
      "grad_norm": 2.4637763500213623,
      "learning_rate": 1e-06,
      "loss": 0.8479,
      "step": 263
    },
    {
      "epoch": 0.03106247793858101,
      "grad_norm": 4.903054237365723,
      "learning_rate": 1e-06,
      "loss": 0.9653,
      "step": 264
    },
    {
      "epoch": 0.03118013883986351,
      "grad_norm": 3.2967870235443115,
      "learning_rate": 1e-06,
      "loss": 0.8837,
      "step": 265
    },
    {
      "epoch": 0.031297799741146014,
      "grad_norm": 2.45072078704834,
      "learning_rate": 1e-06,
      "loss": 0.8762,
      "step": 266
    },
    {
      "epoch": 0.03141546064242852,
      "grad_norm": 20.303709030151367,
      "learning_rate": 1e-06,
      "loss": 0.9166,
      "step": 267
    },
    {
      "epoch": 0.031533121543711025,
      "grad_norm": 6.09748649597168,
      "learning_rate": 1e-06,
      "loss": 0.9721,
      "step": 268
    },
    {
      "epoch": 0.03165078244499353,
      "grad_norm": 2.512051820755005,
      "learning_rate": 1e-06,
      "loss": 0.9464,
      "step": 269
    },
    {
      "epoch": 0.03176844334627603,
      "grad_norm": 4.073144435882568,
      "learning_rate": 1e-06,
      "loss": 0.8717,
      "step": 270
    },
    {
      "epoch": 0.031886104247558535,
      "grad_norm": 4.345306873321533,
      "learning_rate": 1e-06,
      "loss": 0.8799,
      "step": 271
    },
    {
      "epoch": 0.03200376514884104,
      "grad_norm": 4.083920478820801,
      "learning_rate": 1e-06,
      "loss": 1.0133,
      "step": 272
    },
    {
      "epoch": 0.032121426050123546,
      "grad_norm": 2.9541187286376953,
      "learning_rate": 1e-06,
      "loss": 0.9744,
      "step": 273
    },
    {
      "epoch": 0.032239086951406044,
      "grad_norm": 3.141975164413452,
      "learning_rate": 1e-06,
      "loss": 0.9776,
      "step": 274
    },
    {
      "epoch": 0.03235674785268855,
      "grad_norm": 4.691943168640137,
      "learning_rate": 1e-06,
      "loss": 0.883,
      "step": 275
    },
    {
      "epoch": 0.032474408753971055,
      "grad_norm": 5.815479755401611,
      "learning_rate": 1e-06,
      "loss": 0.8965,
      "step": 276
    },
    {
      "epoch": 0.03259206965525356,
      "grad_norm": 4.220132350921631,
      "learning_rate": 1e-06,
      "loss": 0.9603,
      "step": 277
    },
    {
      "epoch": 0.03270973055653606,
      "grad_norm": 3.9638681411743164,
      "learning_rate": 1e-06,
      "loss": 0.9201,
      "step": 278
    },
    {
      "epoch": 0.032827391457818565,
      "grad_norm": 2.6332955360412598,
      "learning_rate": 1e-06,
      "loss": 0.8148,
      "step": 279
    },
    {
      "epoch": 0.03294505235910107,
      "grad_norm": 4.9401397705078125,
      "learning_rate": 1e-06,
      "loss": 0.9787,
      "step": 280
    },
    {
      "epoch": 0.033062713260383576,
      "grad_norm": 3.0980424880981445,
      "learning_rate": 1e-06,
      "loss": 0.8342,
      "step": 281
    },
    {
      "epoch": 0.033180374161666075,
      "grad_norm": 3.9119269847869873,
      "learning_rate": 1e-06,
      "loss": 0.9128,
      "step": 282
    },
    {
      "epoch": 0.03329803506294858,
      "grad_norm": 5.621601104736328,
      "learning_rate": 1e-06,
      "loss": 0.8966,
      "step": 283
    },
    {
      "epoch": 0.033415695964231086,
      "grad_norm": 3.8806064128875732,
      "learning_rate": 1e-06,
      "loss": 0.9079,
      "step": 284
    },
    {
      "epoch": 0.03353335686551359,
      "grad_norm": 4.603435039520264,
      "learning_rate": 1e-06,
      "loss": 0.8963,
      "step": 285
    },
    {
      "epoch": 0.03365101776679609,
      "grad_norm": 4.245962142944336,
      "learning_rate": 1e-06,
      "loss": 0.9149,
      "step": 286
    },
    {
      "epoch": 0.033768678668078596,
      "grad_norm": 4.452668190002441,
      "learning_rate": 1e-06,
      "loss": 0.961,
      "step": 287
    },
    {
      "epoch": 0.0338863395693611,
      "grad_norm": 3.492171287536621,
      "learning_rate": 1e-06,
      "loss": 0.8587,
      "step": 288
    },
    {
      "epoch": 0.03400400047064361,
      "grad_norm": 4.348937511444092,
      "learning_rate": 1e-06,
      "loss": 1.0209,
      "step": 289
    },
    {
      "epoch": 0.034121661371926106,
      "grad_norm": 3.5924530029296875,
      "learning_rate": 1e-06,
      "loss": 0.9734,
      "step": 290
    },
    {
      "epoch": 0.03423932227320861,
      "grad_norm": 4.881322860717773,
      "learning_rate": 1e-06,
      "loss": 0.8807,
      "step": 291
    },
    {
      "epoch": 0.03435698317449112,
      "grad_norm": 3.300433397293091,
      "learning_rate": 1e-06,
      "loss": 0.8877,
      "step": 292
    },
    {
      "epoch": 0.03447464407577362,
      "grad_norm": 5.074859619140625,
      "learning_rate": 1e-06,
      "loss": 0.9591,
      "step": 293
    },
    {
      "epoch": 0.03459230497705612,
      "grad_norm": 4.042046070098877,
      "learning_rate": 1e-06,
      "loss": 0.9333,
      "step": 294
    },
    {
      "epoch": 0.034709965878338626,
      "grad_norm": 5.939486980438232,
      "learning_rate": 1e-06,
      "loss": 0.9156,
      "step": 295
    },
    {
      "epoch": 0.03482762677962113,
      "grad_norm": 9.997541427612305,
      "learning_rate": 1e-06,
      "loss": 0.9161,
      "step": 296
    },
    {
      "epoch": 0.03494528768090364,
      "grad_norm": 3.6703906059265137,
      "learning_rate": 1e-06,
      "loss": 0.8585,
      "step": 297
    },
    {
      "epoch": 0.035062948582186136,
      "grad_norm": 3.217092990875244,
      "learning_rate": 1e-06,
      "loss": 0.9021,
      "step": 298
    },
    {
      "epoch": 0.03518060948346864,
      "grad_norm": 3.009840488433838,
      "learning_rate": 1e-06,
      "loss": 0.9015,
      "step": 299
    },
    {
      "epoch": 0.03529827038475115,
      "grad_norm": 3.5084877014160156,
      "learning_rate": 1e-06,
      "loss": 0.7897,
      "step": 300
    },
    {
      "epoch": 0.03541593128603365,
      "grad_norm": 2.707977056503296,
      "learning_rate": 1e-06,
      "loss": 0.9651,
      "step": 301
    },
    {
      "epoch": 0.03553359218731615,
      "grad_norm": 5.24489164352417,
      "learning_rate": 1e-06,
      "loss": 0.8964,
      "step": 302
    },
    {
      "epoch": 0.03565125308859866,
      "grad_norm": 2.6105916500091553,
      "learning_rate": 1e-06,
      "loss": 0.926,
      "step": 303
    },
    {
      "epoch": 0.03576891398988116,
      "grad_norm": 3.720118761062622,
      "learning_rate": 1e-06,
      "loss": 0.8478,
      "step": 304
    },
    {
      "epoch": 0.03588657489116367,
      "grad_norm": 4.300241470336914,
      "learning_rate": 1e-06,
      "loss": 0.8957,
      "step": 305
    },
    {
      "epoch": 0.03600423579244617,
      "grad_norm": 4.0967512130737305,
      "learning_rate": 1e-06,
      "loss": 0.8142,
      "step": 306
    },
    {
      "epoch": 0.03612189669372867,
      "grad_norm": 4.093652725219727,
      "learning_rate": 1e-06,
      "loss": 0.8621,
      "step": 307
    },
    {
      "epoch": 0.03623955759501118,
      "grad_norm": 3.4667654037475586,
      "learning_rate": 1e-06,
      "loss": 0.8997,
      "step": 308
    },
    {
      "epoch": 0.03635721849629368,
      "grad_norm": 3.124798536300659,
      "learning_rate": 1e-06,
      "loss": 0.8475,
      "step": 309
    },
    {
      "epoch": 0.03647487939757618,
      "grad_norm": 4.9428486824035645,
      "learning_rate": 1e-06,
      "loss": 1.0057,
      "step": 310
    },
    {
      "epoch": 0.03659254029885869,
      "grad_norm": 2.985236167907715,
      "learning_rate": 1e-06,
      "loss": 0.8408,
      "step": 311
    },
    {
      "epoch": 0.03671020120014119,
      "grad_norm": 2.861386775970459,
      "learning_rate": 1e-06,
      "loss": 0.8258,
      "step": 312
    },
    {
      "epoch": 0.0368278621014237,
      "grad_norm": 3.8676459789276123,
      "learning_rate": 1e-06,
      "loss": 0.9179,
      "step": 313
    },
    {
      "epoch": 0.0369455230027062,
      "grad_norm": 4.594061374664307,
      "learning_rate": 1e-06,
      "loss": 0.9291,
      "step": 314
    },
    {
      "epoch": 0.0370631839039887,
      "grad_norm": 3.5613086223602295,
      "learning_rate": 1e-06,
      "loss": 0.9853,
      "step": 315
    },
    {
      "epoch": 0.03718084480527121,
      "grad_norm": 3.170051097869873,
      "learning_rate": 1e-06,
      "loss": 0.8249,
      "step": 316
    },
    {
      "epoch": 0.037298505706553714,
      "grad_norm": 2.451202869415283,
      "learning_rate": 1e-06,
      "loss": 0.9602,
      "step": 317
    },
    {
      "epoch": 0.03741616660783621,
      "grad_norm": 14.758941650390625,
      "learning_rate": 1e-06,
      "loss": 0.7851,
      "step": 318
    },
    {
      "epoch": 0.03753382750911872,
      "grad_norm": 5.861412525177002,
      "learning_rate": 1e-06,
      "loss": 0.9554,
      "step": 319
    },
    {
      "epoch": 0.037651488410401224,
      "grad_norm": 11.434176445007324,
      "learning_rate": 1e-06,
      "loss": 0.8162,
      "step": 320
    },
    {
      "epoch": 0.03776914931168373,
      "grad_norm": 2.5073564052581787,
      "learning_rate": 1e-06,
      "loss": 0.925,
      "step": 321
    },
    {
      "epoch": 0.03788681021296623,
      "grad_norm": 4.871037006378174,
      "learning_rate": 1e-06,
      "loss": 0.9359,
      "step": 322
    },
    {
      "epoch": 0.038004471114248733,
      "grad_norm": 4.676340103149414,
      "learning_rate": 1e-06,
      "loss": 0.8491,
      "step": 323
    },
    {
      "epoch": 0.03812213201553124,
      "grad_norm": 3.828737735748291,
      "learning_rate": 1e-06,
      "loss": 0.8098,
      "step": 324
    },
    {
      "epoch": 0.038239792916813745,
      "grad_norm": 3.128114938735962,
      "learning_rate": 1e-06,
      "loss": 0.8507,
      "step": 325
    },
    {
      "epoch": 0.03835745381809624,
      "grad_norm": 3.4662718772888184,
      "learning_rate": 1e-06,
      "loss": 0.9821,
      "step": 326
    },
    {
      "epoch": 0.03847511471937875,
      "grad_norm": 8.503030776977539,
      "learning_rate": 1e-06,
      "loss": 0.8421,
      "step": 327
    },
    {
      "epoch": 0.038592775620661254,
      "grad_norm": 5.41649055480957,
      "learning_rate": 1e-06,
      "loss": 0.9189,
      "step": 328
    },
    {
      "epoch": 0.03871043652194376,
      "grad_norm": 3.358555793762207,
      "learning_rate": 1e-06,
      "loss": 0.831,
      "step": 329
    },
    {
      "epoch": 0.03882809742322626,
      "grad_norm": 5.377659320831299,
      "learning_rate": 1e-06,
      "loss": 0.8219,
      "step": 330
    },
    {
      "epoch": 0.038945758324508764,
      "grad_norm": 2.828697681427002,
      "learning_rate": 1e-06,
      "loss": 0.9375,
      "step": 331
    },
    {
      "epoch": 0.03906341922579127,
      "grad_norm": 4.016804218292236,
      "learning_rate": 1e-06,
      "loss": 0.9602,
      "step": 332
    },
    {
      "epoch": 0.039181080127073775,
      "grad_norm": 3.2497718334198,
      "learning_rate": 1e-06,
      "loss": 0.8651,
      "step": 333
    },
    {
      "epoch": 0.039298741028356274,
      "grad_norm": 2.9022653102874756,
      "learning_rate": 1e-06,
      "loss": 0.8472,
      "step": 334
    },
    {
      "epoch": 0.03941640192963878,
      "grad_norm": 3.2514617443084717,
      "learning_rate": 1e-06,
      "loss": 1.0098,
      "step": 335
    },
    {
      "epoch": 0.039534062830921285,
      "grad_norm": 5.1763997077941895,
      "learning_rate": 1e-06,
      "loss": 0.8482,
      "step": 336
    },
    {
      "epoch": 0.03965172373220379,
      "grad_norm": 3.454880952835083,
      "learning_rate": 1e-06,
      "loss": 0.9632,
      "step": 337
    },
    {
      "epoch": 0.03976938463348629,
      "grad_norm": 3.575077772140503,
      "learning_rate": 1e-06,
      "loss": 0.8978,
      "step": 338
    },
    {
      "epoch": 0.039887045534768795,
      "grad_norm": 4.193572998046875,
      "learning_rate": 1e-06,
      "loss": 0.9332,
      "step": 339
    },
    {
      "epoch": 0.0400047064360513,
      "grad_norm": 6.426424980163574,
      "learning_rate": 1e-06,
      "loss": 0.8009,
      "step": 340
    },
    {
      "epoch": 0.040122367337333806,
      "grad_norm": 3.2945382595062256,
      "learning_rate": 1e-06,
      "loss": 0.8918,
      "step": 341
    },
    {
      "epoch": 0.040240028238616304,
      "grad_norm": 4.680812358856201,
      "learning_rate": 1e-06,
      "loss": 0.8918,
      "step": 342
    },
    {
      "epoch": 0.04035768913989881,
      "grad_norm": 2.4894278049468994,
      "learning_rate": 1e-06,
      "loss": 0.884,
      "step": 343
    },
    {
      "epoch": 0.040475350041181316,
      "grad_norm": 3.5571224689483643,
      "learning_rate": 1e-06,
      "loss": 0.8372,
      "step": 344
    },
    {
      "epoch": 0.04059301094246382,
      "grad_norm": 5.702209949493408,
      "learning_rate": 1e-06,
      "loss": 0.8881,
      "step": 345
    },
    {
      "epoch": 0.04071067184374632,
      "grad_norm": 4.254310607910156,
      "learning_rate": 1e-06,
      "loss": 0.9545,
      "step": 346
    },
    {
      "epoch": 0.040828332745028825,
      "grad_norm": 5.766552925109863,
      "learning_rate": 1e-06,
      "loss": 0.8308,
      "step": 347
    },
    {
      "epoch": 0.04094599364631133,
      "grad_norm": 5.9502668380737305,
      "learning_rate": 1e-06,
      "loss": 0.863,
      "step": 348
    },
    {
      "epoch": 0.041063654547593836,
      "grad_norm": 3.360926389694214,
      "learning_rate": 1e-06,
      "loss": 0.8732,
      "step": 349
    },
    {
      "epoch": 0.041181315448876335,
      "grad_norm": 7.237725734710693,
      "learning_rate": 1e-06,
      "loss": 0.7984,
      "step": 350
    },
    {
      "epoch": 0.04129897635015884,
      "grad_norm": 2.8876423835754395,
      "learning_rate": 1e-06,
      "loss": 0.875,
      "step": 351
    },
    {
      "epoch": 0.041416637251441346,
      "grad_norm": 3.3526382446289062,
      "learning_rate": 1e-06,
      "loss": 0.8522,
      "step": 352
    },
    {
      "epoch": 0.04153429815272385,
      "grad_norm": 14.322229385375977,
      "learning_rate": 1e-06,
      "loss": 0.7397,
      "step": 353
    },
    {
      "epoch": 0.04165195905400635,
      "grad_norm": 4.687523365020752,
      "learning_rate": 1e-06,
      "loss": 0.7707,
      "step": 354
    },
    {
      "epoch": 0.041769619955288856,
      "grad_norm": 3.8648178577423096,
      "learning_rate": 1e-06,
      "loss": 0.812,
      "step": 355
    },
    {
      "epoch": 0.04188728085657136,
      "grad_norm": 4.16708517074585,
      "learning_rate": 1e-06,
      "loss": 0.8107,
      "step": 356
    },
    {
      "epoch": 0.04200494175785387,
      "grad_norm": 5.837557792663574,
      "learning_rate": 1e-06,
      "loss": 0.7445,
      "step": 357
    },
    {
      "epoch": 0.042122602659136366,
      "grad_norm": 3.4978277683258057,
      "learning_rate": 1e-06,
      "loss": 0.9116,
      "step": 358
    },
    {
      "epoch": 0.04224026356041887,
      "grad_norm": 2.914486885070801,
      "learning_rate": 1e-06,
      "loss": 0.9159,
      "step": 359
    },
    {
      "epoch": 0.04235792446170138,
      "grad_norm": 3.2980706691741943,
      "learning_rate": 1e-06,
      "loss": 0.8705,
      "step": 360
    },
    {
      "epoch": 0.04247558536298388,
      "grad_norm": 6.628724098205566,
      "learning_rate": 1e-06,
      "loss": 0.861,
      "step": 361
    },
    {
      "epoch": 0.04259324626426638,
      "grad_norm": 3.676485776901245,
      "learning_rate": 1e-06,
      "loss": 0.7846,
      "step": 362
    },
    {
      "epoch": 0.042710907165548886,
      "grad_norm": 5.055513381958008,
      "learning_rate": 1e-06,
      "loss": 0.7648,
      "step": 363
    },
    {
      "epoch": 0.04282856806683139,
      "grad_norm": 4.655031681060791,
      "learning_rate": 1e-06,
      "loss": 0.8165,
      "step": 364
    },
    {
      "epoch": 0.0429462289681139,
      "grad_norm": 6.790380954742432,
      "learning_rate": 1e-06,
      "loss": 0.8033,
      "step": 365
    },
    {
      "epoch": 0.043063889869396396,
      "grad_norm": 3.321683645248413,
      "learning_rate": 1e-06,
      "loss": 0.8015,
      "step": 366
    },
    {
      "epoch": 0.0431815507706789,
      "grad_norm": 6.742896556854248,
      "learning_rate": 1e-06,
      "loss": 0.8881,
      "step": 367
    },
    {
      "epoch": 0.04329921167196141,
      "grad_norm": 4.620999336242676,
      "learning_rate": 1e-06,
      "loss": 0.8187,
      "step": 368
    },
    {
      "epoch": 0.04341687257324391,
      "grad_norm": 3.710247755050659,
      "learning_rate": 1e-06,
      "loss": 1.1444,
      "step": 369
    },
    {
      "epoch": 0.04353453347452641,
      "grad_norm": 4.3619866371154785,
      "learning_rate": 1e-06,
      "loss": 0.8836,
      "step": 370
    },
    {
      "epoch": 0.04365219437580892,
      "grad_norm": 4.016956806182861,
      "learning_rate": 1e-06,
      "loss": 0.8193,
      "step": 371
    },
    {
      "epoch": 0.04376985527709142,
      "grad_norm": 3.5968549251556396,
      "learning_rate": 1e-06,
      "loss": 0.8534,
      "step": 372
    },
    {
      "epoch": 0.04388751617837393,
      "grad_norm": 5.6325836181640625,
      "learning_rate": 1e-06,
      "loss": 0.7794,
      "step": 373
    },
    {
      "epoch": 0.04400517707965643,
      "grad_norm": 5.343469142913818,
      "learning_rate": 1e-06,
      "loss": 0.876,
      "step": 374
    },
    {
      "epoch": 0.04412283798093893,
      "grad_norm": 3.2438323497772217,
      "learning_rate": 1e-06,
      "loss": 0.809,
      "step": 375
    },
    {
      "epoch": 0.04424049888222144,
      "grad_norm": 3.4838759899139404,
      "learning_rate": 1e-06,
      "loss": 0.8749,
      "step": 376
    },
    {
      "epoch": 0.04435815978350394,
      "grad_norm": 3.5921363830566406,
      "learning_rate": 1e-06,
      "loss": 0.8455,
      "step": 377
    },
    {
      "epoch": 0.04447582068478644,
      "grad_norm": 2.481532335281372,
      "learning_rate": 1e-06,
      "loss": 0.9034,
      "step": 378
    },
    {
      "epoch": 0.04459348158606895,
      "grad_norm": 2.756016731262207,
      "learning_rate": 1e-06,
      "loss": 0.8722,
      "step": 379
    },
    {
      "epoch": 0.04471114248735145,
      "grad_norm": 4.9279913902282715,
      "learning_rate": 1e-06,
      "loss": 0.7027,
      "step": 380
    },
    {
      "epoch": 0.04482880338863396,
      "grad_norm": 4.436341285705566,
      "learning_rate": 1e-06,
      "loss": 0.9594,
      "step": 381
    },
    {
      "epoch": 0.04494646428991646,
      "grad_norm": 6.906442165374756,
      "learning_rate": 1e-06,
      "loss": 0.8929,
      "step": 382
    },
    {
      "epoch": 0.04506412519119896,
      "grad_norm": 3.315844774246216,
      "learning_rate": 1e-06,
      "loss": 0.7651,
      "step": 383
    },
    {
      "epoch": 0.04518178609248147,
      "grad_norm": 5.387514591217041,
      "learning_rate": 1e-06,
      "loss": 0.826,
      "step": 384
    },
    {
      "epoch": 0.045299446993763974,
      "grad_norm": 4.287894248962402,
      "learning_rate": 1e-06,
      "loss": 0.8288,
      "step": 385
    },
    {
      "epoch": 0.04541710789504647,
      "grad_norm": 3.736056327819824,
      "learning_rate": 1e-06,
      "loss": 0.9377,
      "step": 386
    },
    {
      "epoch": 0.04553476879632898,
      "grad_norm": 3.92673659324646,
      "learning_rate": 1e-06,
      "loss": 0.8998,
      "step": 387
    },
    {
      "epoch": 0.045652429697611484,
      "grad_norm": 6.745547294616699,
      "learning_rate": 1e-06,
      "loss": 0.8782,
      "step": 388
    },
    {
      "epoch": 0.04577009059889399,
      "grad_norm": 3.430509567260742,
      "learning_rate": 1e-06,
      "loss": 0.8118,
      "step": 389
    },
    {
      "epoch": 0.04588775150017649,
      "grad_norm": 3.780449867248535,
      "learning_rate": 1e-06,
      "loss": 0.7847,
      "step": 390
    },
    {
      "epoch": 0.046005412401458994,
      "grad_norm": 5.1995320320129395,
      "learning_rate": 1e-06,
      "loss": 0.7804,
      "step": 391
    },
    {
      "epoch": 0.0461230733027415,
      "grad_norm": 4.746626377105713,
      "learning_rate": 1e-06,
      "loss": 0.8891,
      "step": 392
    },
    {
      "epoch": 0.046240734204024005,
      "grad_norm": 3.883009433746338,
      "learning_rate": 1e-06,
      "loss": 0.8285,
      "step": 393
    },
    {
      "epoch": 0.0463583951053065,
      "grad_norm": 3.148414134979248,
      "learning_rate": 1e-06,
      "loss": 0.8413,
      "step": 394
    },
    {
      "epoch": 0.04647605600658901,
      "grad_norm": 3.574962854385376,
      "learning_rate": 1e-06,
      "loss": 0.8589,
      "step": 395
    },
    {
      "epoch": 0.046593716907871514,
      "grad_norm": 3.7145674228668213,
      "learning_rate": 1e-06,
      "loss": 0.8848,
      "step": 396
    },
    {
      "epoch": 0.04671137780915402,
      "grad_norm": 2.2142579555511475,
      "learning_rate": 1e-06,
      "loss": 0.9702,
      "step": 397
    },
    {
      "epoch": 0.04682903871043652,
      "grad_norm": 3.7105839252471924,
      "learning_rate": 1e-06,
      "loss": 0.8114,
      "step": 398
    },
    {
      "epoch": 0.046946699611719024,
      "grad_norm": 3.403414249420166,
      "learning_rate": 1e-06,
      "loss": 0.9485,
      "step": 399
    },
    {
      "epoch": 0.04706436051300153,
      "grad_norm": 5.079216957092285,
      "learning_rate": 1e-06,
      "loss": 0.8575,
      "step": 400
    },
    {
      "epoch": 0.047182021414284035,
      "grad_norm": 5.360959529876709,
      "learning_rate": 1e-06,
      "loss": 0.9164,
      "step": 401
    },
    {
      "epoch": 0.047299682315566534,
      "grad_norm": 2.644434928894043,
      "learning_rate": 1e-06,
      "loss": 0.9584,
      "step": 402
    },
    {
      "epoch": 0.04741734321684904,
      "grad_norm": 3.6469929218292236,
      "learning_rate": 1e-06,
      "loss": 0.9308,
      "step": 403
    },
    {
      "epoch": 0.047535004118131545,
      "grad_norm": 4.576849460601807,
      "learning_rate": 1e-06,
      "loss": 0.9586,
      "step": 404
    },
    {
      "epoch": 0.04765266501941405,
      "grad_norm": 3.8551862239837646,
      "learning_rate": 1e-06,
      "loss": 0.859,
      "step": 405
    },
    {
      "epoch": 0.04777032592069655,
      "grad_norm": 6.5358452796936035,
      "learning_rate": 1e-06,
      "loss": 0.8241,
      "step": 406
    },
    {
      "epoch": 0.047887986821979055,
      "grad_norm": 2.817338466644287,
      "learning_rate": 1e-06,
      "loss": 0.8393,
      "step": 407
    },
    {
      "epoch": 0.04800564772326156,
      "grad_norm": 2.9196722507476807,
      "learning_rate": 1e-06,
      "loss": 0.8141,
      "step": 408
    },
    {
      "epoch": 0.048123308624544066,
      "grad_norm": 5.13906717300415,
      "learning_rate": 1e-06,
      "loss": 0.8182,
      "step": 409
    },
    {
      "epoch": 0.048240969525826564,
      "grad_norm": 3.631796360015869,
      "learning_rate": 1e-06,
      "loss": 0.8274,
      "step": 410
    },
    {
      "epoch": 0.04835863042710907,
      "grad_norm": 4.240082263946533,
      "learning_rate": 1e-06,
      "loss": 0.8955,
      "step": 411
    },
    {
      "epoch": 0.048476291328391576,
      "grad_norm": 3.6155200004577637,
      "learning_rate": 1e-06,
      "loss": 0.8182,
      "step": 412
    },
    {
      "epoch": 0.04859395222967408,
      "grad_norm": 5.879213333129883,
      "learning_rate": 1e-06,
      "loss": 0.8166,
      "step": 413
    },
    {
      "epoch": 0.04871161313095658,
      "grad_norm": 6.984928131103516,
      "learning_rate": 1e-06,
      "loss": 0.8011,
      "step": 414
    },
    {
      "epoch": 0.048829274032239085,
      "grad_norm": 8.660188674926758,
      "learning_rate": 1e-06,
      "loss": 0.807,
      "step": 415
    },
    {
      "epoch": 0.04894693493352159,
      "grad_norm": 5.219287395477295,
      "learning_rate": 1e-06,
      "loss": 0.7783,
      "step": 416
    },
    {
      "epoch": 0.049064595834804096,
      "grad_norm": 3.9182770252227783,
      "learning_rate": 1e-06,
      "loss": 0.7483,
      "step": 417
    },
    {
      "epoch": 0.049182256736086595,
      "grad_norm": 5.254255294799805,
      "learning_rate": 1e-06,
      "loss": 0.819,
      "step": 418
    },
    {
      "epoch": 0.0492999176373691,
      "grad_norm": 5.512103080749512,
      "learning_rate": 1e-06,
      "loss": 1.0628,
      "step": 419
    },
    {
      "epoch": 0.049417578538651606,
      "grad_norm": 5.233940601348877,
      "learning_rate": 1e-06,
      "loss": 0.8147,
      "step": 420
    },
    {
      "epoch": 0.04953523943993411,
      "grad_norm": 4.2004547119140625,
      "learning_rate": 1e-06,
      "loss": 0.8236,
      "step": 421
    },
    {
      "epoch": 0.04965290034121661,
      "grad_norm": 3.2189583778381348,
      "learning_rate": 1e-06,
      "loss": 0.8067,
      "step": 422
    },
    {
      "epoch": 0.049770561242499116,
      "grad_norm": 3.050485610961914,
      "learning_rate": 1e-06,
      "loss": 0.875,
      "step": 423
    },
    {
      "epoch": 0.04988822214378162,
      "grad_norm": 5.160003662109375,
      "learning_rate": 1e-06,
      "loss": 0.8419,
      "step": 424
    },
    {
      "epoch": 0.05000588304506413,
      "grad_norm": 7.041035175323486,
      "learning_rate": 1e-06,
      "loss": 0.8879,
      "step": 425
    },
    {
      "epoch": 0.050123543946346626,
      "grad_norm": 4.338962078094482,
      "learning_rate": 1e-06,
      "loss": 0.8643,
      "step": 426
    },
    {
      "epoch": 0.05024120484762913,
      "grad_norm": 5.213710308074951,
      "learning_rate": 1e-06,
      "loss": 0.7715,
      "step": 427
    },
    {
      "epoch": 0.05035886574891164,
      "grad_norm": 2.2304441928863525,
      "learning_rate": 1e-06,
      "loss": 0.8655,
      "step": 428
    },
    {
      "epoch": 0.05047652665019414,
      "grad_norm": 3.0590453147888184,
      "learning_rate": 1e-06,
      "loss": 0.9235,
      "step": 429
    },
    {
      "epoch": 0.05059418755147664,
      "grad_norm": 5.010176181793213,
      "learning_rate": 1e-06,
      "loss": 0.8031,
      "step": 430
    },
    {
      "epoch": 0.050711848452759146,
      "grad_norm": 2.6110637187957764,
      "learning_rate": 1e-06,
      "loss": 0.8778,
      "step": 431
    },
    {
      "epoch": 0.05082950935404165,
      "grad_norm": 3.213735818862915,
      "learning_rate": 1e-06,
      "loss": 0.7609,
      "step": 432
    },
    {
      "epoch": 0.05094717025532416,
      "grad_norm": 3.956066131591797,
      "learning_rate": 1e-06,
      "loss": 0.7591,
      "step": 433
    },
    {
      "epoch": 0.051064831156606656,
      "grad_norm": 21.855979919433594,
      "learning_rate": 1e-06,
      "loss": 0.8767,
      "step": 434
    },
    {
      "epoch": 0.05118249205788916,
      "grad_norm": 9.275443077087402,
      "learning_rate": 1e-06,
      "loss": 0.7895,
      "step": 435
    },
    {
      "epoch": 0.05130015295917167,
      "grad_norm": 3.2387189865112305,
      "learning_rate": 1e-06,
      "loss": 0.8595,
      "step": 436
    },
    {
      "epoch": 0.05141781386045417,
      "grad_norm": 2.4975271224975586,
      "learning_rate": 1e-06,
      "loss": 0.773,
      "step": 437
    },
    {
      "epoch": 0.05153547476173667,
      "grad_norm": 7.902890205383301,
      "learning_rate": 1e-06,
      "loss": 0.7657,
      "step": 438
    },
    {
      "epoch": 0.05165313566301918,
      "grad_norm": 3.4846811294555664,
      "learning_rate": 1e-06,
      "loss": 0.9463,
      "step": 439
    },
    {
      "epoch": 0.05177079656430168,
      "grad_norm": 12.962797164916992,
      "learning_rate": 1e-06,
      "loss": 0.7384,
      "step": 440
    },
    {
      "epoch": 0.05188845746558419,
      "grad_norm": 2.69588041305542,
      "learning_rate": 1e-06,
      "loss": 0.7672,
      "step": 441
    },
    {
      "epoch": 0.05200611836686669,
      "grad_norm": 2.473055839538574,
      "learning_rate": 1e-06,
      "loss": 0.8804,
      "step": 442
    },
    {
      "epoch": 0.05212377926814919,
      "grad_norm": 4.784314155578613,
      "learning_rate": 1e-06,
      "loss": 0.7457,
      "step": 443
    },
    {
      "epoch": 0.0522414401694317,
      "grad_norm": 3.915424346923828,
      "learning_rate": 1e-06,
      "loss": 0.8125,
      "step": 444
    },
    {
      "epoch": 0.052359101070714203,
      "grad_norm": 3.773595094680786,
      "learning_rate": 1e-06,
      "loss": 0.8424,
      "step": 445
    },
    {
      "epoch": 0.0524767619719967,
      "grad_norm": 4.634319305419922,
      "learning_rate": 1e-06,
      "loss": 0.7904,
      "step": 446
    },
    {
      "epoch": 0.05259442287327921,
      "grad_norm": 4.115906715393066,
      "learning_rate": 1e-06,
      "loss": 0.7609,
      "step": 447
    },
    {
      "epoch": 0.05271208377456171,
      "grad_norm": 8.289746284484863,
      "learning_rate": 1e-06,
      "loss": 0.8442,
      "step": 448
    },
    {
      "epoch": 0.05282974467584422,
      "grad_norm": 21.677457809448242,
      "learning_rate": 1e-06,
      "loss": 0.8504,
      "step": 449
    },
    {
      "epoch": 0.05294740557712672,
      "grad_norm": 4.526599884033203,
      "learning_rate": 1e-06,
      "loss": 0.8961,
      "step": 450
    },
    {
      "epoch": 0.05306506647840922,
      "grad_norm": 6.855340003967285,
      "learning_rate": 1e-06,
      "loss": 0.8506,
      "step": 451
    },
    {
      "epoch": 0.05318272737969173,
      "grad_norm": 4.277021884918213,
      "learning_rate": 1e-06,
      "loss": 0.7808,
      "step": 452
    },
    {
      "epoch": 0.053300388280974234,
      "grad_norm": 8.101177215576172,
      "learning_rate": 1e-06,
      "loss": 0.7786,
      "step": 453
    },
    {
      "epoch": 0.05341804918225673,
      "grad_norm": 2.057940721511841,
      "learning_rate": 1e-06,
      "loss": 0.8495,
      "step": 454
    },
    {
      "epoch": 0.05353571008353924,
      "grad_norm": 4.969939708709717,
      "learning_rate": 1e-06,
      "loss": 0.7586,
      "step": 455
    },
    {
      "epoch": 0.053653370984821744,
      "grad_norm": 2.957061290740967,
      "learning_rate": 1e-06,
      "loss": 0.8525,
      "step": 456
    },
    {
      "epoch": 0.05377103188610425,
      "grad_norm": 8.066761016845703,
      "learning_rate": 1e-06,
      "loss": 0.7433,
      "step": 457
    },
    {
      "epoch": 0.05388869278738675,
      "grad_norm": 4.397155284881592,
      "learning_rate": 1e-06,
      "loss": 0.7453,
      "step": 458
    },
    {
      "epoch": 0.054006353688669254,
      "grad_norm": 3.529214382171631,
      "learning_rate": 1e-06,
      "loss": 0.9033,
      "step": 459
    },
    {
      "epoch": 0.05412401458995176,
      "grad_norm": 5.184650421142578,
      "learning_rate": 1e-06,
      "loss": 0.9159,
      "step": 460
    },
    {
      "epoch": 0.054241675491234265,
      "grad_norm": 3.8979060649871826,
      "learning_rate": 1e-06,
      "loss": 0.9171,
      "step": 461
    },
    {
      "epoch": 0.05435933639251676,
      "grad_norm": 3.0818674564361572,
      "learning_rate": 1e-06,
      "loss": 0.8038,
      "step": 462
    },
    {
      "epoch": 0.05447699729379927,
      "grad_norm": 4.317335605621338,
      "learning_rate": 1e-06,
      "loss": 0.7565,
      "step": 463
    },
    {
      "epoch": 0.054594658195081774,
      "grad_norm": 3.5969228744506836,
      "learning_rate": 1e-06,
      "loss": 0.7188,
      "step": 464
    },
    {
      "epoch": 0.05471231909636428,
      "grad_norm": 4.5173115730285645,
      "learning_rate": 1e-06,
      "loss": 0.8674,
      "step": 465
    },
    {
      "epoch": 0.05482997999764678,
      "grad_norm": 3.263984441757202,
      "learning_rate": 1e-06,
      "loss": 0.8585,
      "step": 466
    },
    {
      "epoch": 0.054947640898929284,
      "grad_norm": 3.5642848014831543,
      "learning_rate": 1e-06,
      "loss": 0.8787,
      "step": 467
    },
    {
      "epoch": 0.05506530180021179,
      "grad_norm": 5.6151347160339355,
      "learning_rate": 1e-06,
      "loss": 0.7276,
      "step": 468
    },
    {
      "epoch": 0.055182962701494295,
      "grad_norm": 3.163313150405884,
      "learning_rate": 1e-06,
      "loss": 0.7931,
      "step": 469
    },
    {
      "epoch": 0.055300623602776794,
      "grad_norm": 2.878931999206543,
      "learning_rate": 1e-06,
      "loss": 0.7667,
      "step": 470
    },
    {
      "epoch": 0.0554182845040593,
      "grad_norm": 3.318117618560791,
      "learning_rate": 1e-06,
      "loss": 0.7391,
      "step": 471
    },
    {
      "epoch": 0.055535945405341805,
      "grad_norm": 3.127615213394165,
      "learning_rate": 1e-06,
      "loss": 0.9311,
      "step": 472
    },
    {
      "epoch": 0.05565360630662431,
      "grad_norm": 2.90480637550354,
      "learning_rate": 1e-06,
      "loss": 0.9328,
      "step": 473
    },
    {
      "epoch": 0.05577126720790681,
      "grad_norm": 3.7865421772003174,
      "learning_rate": 1e-06,
      "loss": 0.745,
      "step": 474
    },
    {
      "epoch": 0.055888928109189315,
      "grad_norm": 3.2640819549560547,
      "learning_rate": 1e-06,
      "loss": 0.9306,
      "step": 475
    },
    {
      "epoch": 0.05600658901047182,
      "grad_norm": 3.227083206176758,
      "learning_rate": 1e-06,
      "loss": 0.8744,
      "step": 476
    },
    {
      "epoch": 0.056124249911754326,
      "grad_norm": 3.939415454864502,
      "learning_rate": 1e-06,
      "loss": 0.8885,
      "step": 477
    },
    {
      "epoch": 0.056241910813036824,
      "grad_norm": 2.41015887260437,
      "learning_rate": 1e-06,
      "loss": 0.8947,
      "step": 478
    },
    {
      "epoch": 0.05635957171431933,
      "grad_norm": 4.985194206237793,
      "learning_rate": 1e-06,
      "loss": 0.8153,
      "step": 479
    },
    {
      "epoch": 0.056477232615601836,
      "grad_norm": 3.3147754669189453,
      "learning_rate": 1e-06,
      "loss": 0.8851,
      "step": 480
    },
    {
      "epoch": 0.05659489351688434,
      "grad_norm": 6.035532474517822,
      "learning_rate": 1e-06,
      "loss": 0.8393,
      "step": 481
    },
    {
      "epoch": 0.05671255441816684,
      "grad_norm": 5.8169403076171875,
      "learning_rate": 1e-06,
      "loss": 0.8692,
      "step": 482
    },
    {
      "epoch": 0.056830215319449345,
      "grad_norm": 3.139738082885742,
      "learning_rate": 1e-06,
      "loss": 0.8567,
      "step": 483
    },
    {
      "epoch": 0.05694787622073185,
      "grad_norm": 6.8116021156311035,
      "learning_rate": 1e-06,
      "loss": 0.8185,
      "step": 484
    },
    {
      "epoch": 0.057065537122014356,
      "grad_norm": 4.787223815917969,
      "learning_rate": 1e-06,
      "loss": 0.726,
      "step": 485
    },
    {
      "epoch": 0.057183198023296855,
      "grad_norm": 2.7163264751434326,
      "learning_rate": 1e-06,
      "loss": 0.8637,
      "step": 486
    },
    {
      "epoch": 0.05730085892457936,
      "grad_norm": 4.378730773925781,
      "learning_rate": 1e-06,
      "loss": 0.9133,
      "step": 487
    },
    {
      "epoch": 0.057418519825861866,
      "grad_norm": 4.203522682189941,
      "learning_rate": 1e-06,
      "loss": 0.8322,
      "step": 488
    },
    {
      "epoch": 0.05753618072714437,
      "grad_norm": 6.0501837730407715,
      "learning_rate": 1e-06,
      "loss": 0.9354,
      "step": 489
    },
    {
      "epoch": 0.05765384162842687,
      "grad_norm": 4.329912185668945,
      "learning_rate": 1e-06,
      "loss": 0.6985,
      "step": 490
    },
    {
      "epoch": 0.057771502529709376,
      "grad_norm": 16.60413360595703,
      "learning_rate": 1e-06,
      "loss": 0.7859,
      "step": 491
    },
    {
      "epoch": 0.05788916343099188,
      "grad_norm": 4.150776386260986,
      "learning_rate": 1e-06,
      "loss": 0.8087,
      "step": 492
    },
    {
      "epoch": 0.05800682433227439,
      "grad_norm": 4.654838562011719,
      "learning_rate": 1e-06,
      "loss": 0.8132,
      "step": 493
    },
    {
      "epoch": 0.058124485233556886,
      "grad_norm": 6.413145542144775,
      "learning_rate": 1e-06,
      "loss": 0.7564,
      "step": 494
    },
    {
      "epoch": 0.05824214613483939,
      "grad_norm": 4.651871204376221,
      "learning_rate": 1e-06,
      "loss": 0.8122,
      "step": 495
    },
    {
      "epoch": 0.0583598070361219,
      "grad_norm": 4.54334020614624,
      "learning_rate": 1e-06,
      "loss": 0.7293,
      "step": 496
    },
    {
      "epoch": 0.0584774679374044,
      "grad_norm": 2.9437007904052734,
      "learning_rate": 1e-06,
      "loss": 0.8403,
      "step": 497
    },
    {
      "epoch": 0.0585951288386869,
      "grad_norm": 2.51481294631958,
      "learning_rate": 1e-06,
      "loss": 0.7564,
      "step": 498
    },
    {
      "epoch": 0.05871278973996941,
      "grad_norm": 3.907083749771118,
      "learning_rate": 1e-06,
      "loss": 0.8362,
      "step": 499
    },
    {
      "epoch": 0.05883045064125191,
      "grad_norm": 2.8982291221618652,
      "learning_rate": 1e-06,
      "loss": 0.7787,
      "step": 500
    },
    {
      "epoch": 0.05894811154253442,
      "grad_norm": 4.202147960662842,
      "learning_rate": 1e-06,
      "loss": 0.7667,
      "step": 501
    },
    {
      "epoch": 0.059065772443816916,
      "grad_norm": 2.803162097930908,
      "learning_rate": 1e-06,
      "loss": 0.7728,
      "step": 502
    },
    {
      "epoch": 0.05918343334509942,
      "grad_norm": 2.6605453491210938,
      "learning_rate": 1e-06,
      "loss": 0.8435,
      "step": 503
    },
    {
      "epoch": 0.05930109424638193,
      "grad_norm": 3.9412362575531006,
      "learning_rate": 1e-06,
      "loss": 1.0537,
      "step": 504
    },
    {
      "epoch": 0.05941875514766443,
      "grad_norm": 4.297357082366943,
      "learning_rate": 1e-06,
      "loss": 0.7846,
      "step": 505
    },
    {
      "epoch": 0.05953641604894693,
      "grad_norm": 3.5739941596984863,
      "learning_rate": 1e-06,
      "loss": 0.9041,
      "step": 506
    },
    {
      "epoch": 0.05965407695022944,
      "grad_norm": 4.355228900909424,
      "learning_rate": 1e-06,
      "loss": 0.7947,
      "step": 507
    },
    {
      "epoch": 0.05977173785151194,
      "grad_norm": 3.83013916015625,
      "learning_rate": 1e-06,
      "loss": 0.8155,
      "step": 508
    },
    {
      "epoch": 0.05988939875279445,
      "grad_norm": 4.5131049156188965,
      "learning_rate": 1e-06,
      "loss": 0.8494,
      "step": 509
    },
    {
      "epoch": 0.06000705965407695,
      "grad_norm": 2.556640148162842,
      "learning_rate": 1e-06,
      "loss": 0.8655,
      "step": 510
    },
    {
      "epoch": 0.06012472055535945,
      "grad_norm": 4.710814952850342,
      "learning_rate": 1e-06,
      "loss": 0.8462,
      "step": 511
    },
    {
      "epoch": 0.06024238145664196,
      "grad_norm": 2.3375051021575928,
      "learning_rate": 1e-06,
      "loss": 0.8662,
      "step": 512
    },
    {
      "epoch": 0.060360042357924464,
      "grad_norm": 6.374229431152344,
      "learning_rate": 1e-06,
      "loss": 0.8735,
      "step": 513
    },
    {
      "epoch": 0.06047770325920696,
      "grad_norm": 4.315854072570801,
      "learning_rate": 1e-06,
      "loss": 0.7404,
      "step": 514
    },
    {
      "epoch": 0.06059536416048947,
      "grad_norm": 7.783134937286377,
      "learning_rate": 1e-06,
      "loss": 0.79,
      "step": 515
    },
    {
      "epoch": 0.06071302506177197,
      "grad_norm": 2.5863089561462402,
      "learning_rate": 1e-06,
      "loss": 0.9462,
      "step": 516
    },
    {
      "epoch": 0.06083068596305448,
      "grad_norm": 8.600770950317383,
      "learning_rate": 1e-06,
      "loss": 0.8503,
      "step": 517
    },
    {
      "epoch": 0.06094834686433698,
      "grad_norm": 6.381835460662842,
      "learning_rate": 1e-06,
      "loss": 0.8524,
      "step": 518
    },
    {
      "epoch": 0.06106600776561948,
      "grad_norm": 9.072025299072266,
      "learning_rate": 1e-06,
      "loss": 0.8341,
      "step": 519
    },
    {
      "epoch": 0.06118366866690199,
      "grad_norm": 4.649893283843994,
      "learning_rate": 1e-06,
      "loss": 0.8017,
      "step": 520
    },
    {
      "epoch": 0.061301329568184494,
      "grad_norm": 3.5497517585754395,
      "learning_rate": 1e-06,
      "loss": 0.822,
      "step": 521
    },
    {
      "epoch": 0.06141899046946699,
      "grad_norm": 6.476340293884277,
      "learning_rate": 1e-06,
      "loss": 0.7008,
      "step": 522
    },
    {
      "epoch": 0.0615366513707495,
      "grad_norm": 4.787139892578125,
      "learning_rate": 1e-06,
      "loss": 0.682,
      "step": 523
    },
    {
      "epoch": 0.061654312272032004,
      "grad_norm": 3.200488328933716,
      "learning_rate": 1e-06,
      "loss": 0.8403,
      "step": 524
    },
    {
      "epoch": 0.06177197317331451,
      "grad_norm": 4.321343898773193,
      "learning_rate": 1e-06,
      "loss": 0.713,
      "step": 525
    },
    {
      "epoch": 0.06188963407459701,
      "grad_norm": 3.959671974182129,
      "learning_rate": 1e-06,
      "loss": 0.7889,
      "step": 526
    },
    {
      "epoch": 0.062007294975879514,
      "grad_norm": 2.6865928173065186,
      "learning_rate": 1e-06,
      "loss": 0.7698,
      "step": 527
    },
    {
      "epoch": 0.06212495587716202,
      "grad_norm": 4.1267170906066895,
      "learning_rate": 1e-06,
      "loss": 0.9351,
      "step": 528
    },
    {
      "epoch": 0.062242616778444525,
      "grad_norm": 3.508446216583252,
      "learning_rate": 1e-06,
      "loss": 0.7834,
      "step": 529
    },
    {
      "epoch": 0.06236027767972702,
      "grad_norm": 11.811966896057129,
      "learning_rate": 1e-06,
      "loss": 0.6957,
      "step": 530
    },
    {
      "epoch": 0.06247793858100953,
      "grad_norm": 3.9337544441223145,
      "learning_rate": 1e-06,
      "loss": 0.9831,
      "step": 531
    },
    {
      "epoch": 0.06259559948229203,
      "grad_norm": 5.258102893829346,
      "learning_rate": 1e-06,
      "loss": 0.8131,
      "step": 532
    },
    {
      "epoch": 0.06271326038357454,
      "grad_norm": 3.2443442344665527,
      "learning_rate": 1e-06,
      "loss": 0.8168,
      "step": 533
    },
    {
      "epoch": 0.06283092128485704,
      "grad_norm": 7.886223316192627,
      "learning_rate": 1e-06,
      "loss": 0.9043,
      "step": 534
    },
    {
      "epoch": 0.06294858218613955,
      "grad_norm": 2.490288257598877,
      "learning_rate": 1e-06,
      "loss": 0.8503,
      "step": 535
    },
    {
      "epoch": 0.06306624308742205,
      "grad_norm": 9.365945816040039,
      "learning_rate": 1e-06,
      "loss": 0.8139,
      "step": 536
    },
    {
      "epoch": 0.06318390398870455,
      "grad_norm": 3.3306145668029785,
      "learning_rate": 1e-06,
      "loss": 0.8527,
      "step": 537
    },
    {
      "epoch": 0.06330156488998706,
      "grad_norm": 4.600078105926514,
      "learning_rate": 1e-06,
      "loss": 0.7503,
      "step": 538
    },
    {
      "epoch": 0.06341922579126956,
      "grad_norm": 2.8406596183776855,
      "learning_rate": 1e-06,
      "loss": 0.7695,
      "step": 539
    },
    {
      "epoch": 0.06353688669255206,
      "grad_norm": 10.398956298828125,
      "learning_rate": 1e-06,
      "loss": 0.7327,
      "step": 540
    },
    {
      "epoch": 0.06365454759383457,
      "grad_norm": 6.3080735206604,
      "learning_rate": 1e-06,
      "loss": 0.7773,
      "step": 541
    },
    {
      "epoch": 0.06377220849511707,
      "grad_norm": 4.205158710479736,
      "learning_rate": 1e-06,
      "loss": 0.8143,
      "step": 542
    },
    {
      "epoch": 0.06388986939639958,
      "grad_norm": 4.888593673706055,
      "learning_rate": 1e-06,
      "loss": 0.795,
      "step": 543
    },
    {
      "epoch": 0.06400753029768208,
      "grad_norm": 3.363938570022583,
      "learning_rate": 1e-06,
      "loss": 0.7612,
      "step": 544
    },
    {
      "epoch": 0.06412519119896458,
      "grad_norm": 3.169189691543579,
      "learning_rate": 1e-06,
      "loss": 0.8616,
      "step": 545
    },
    {
      "epoch": 0.06424285210024709,
      "grad_norm": 4.985038757324219,
      "learning_rate": 1e-06,
      "loss": 0.8326,
      "step": 546
    },
    {
      "epoch": 0.06436051300152959,
      "grad_norm": 4.971280574798584,
      "learning_rate": 1e-06,
      "loss": 0.806,
      "step": 547
    },
    {
      "epoch": 0.06447817390281209,
      "grad_norm": 5.460928440093994,
      "learning_rate": 1e-06,
      "loss": 0.7662,
      "step": 548
    },
    {
      "epoch": 0.0645958348040946,
      "grad_norm": 4.207946300506592,
      "learning_rate": 1e-06,
      "loss": 0.6787,
      "step": 549
    },
    {
      "epoch": 0.0647134957053771,
      "grad_norm": 2.679123640060425,
      "learning_rate": 1e-06,
      "loss": 0.7729,
      "step": 550
    },
    {
      "epoch": 0.06483115660665961,
      "grad_norm": 3.993603467941284,
      "learning_rate": 1e-06,
      "loss": 0.8018,
      "step": 551
    },
    {
      "epoch": 0.06494881750794211,
      "grad_norm": 4.020260810852051,
      "learning_rate": 1e-06,
      "loss": 0.7647,
      "step": 552
    },
    {
      "epoch": 0.06506647840922461,
      "grad_norm": 2.9653048515319824,
      "learning_rate": 1e-06,
      "loss": 0.7791,
      "step": 553
    },
    {
      "epoch": 0.06518413931050712,
      "grad_norm": 2.8370585441589355,
      "learning_rate": 1e-06,
      "loss": 0.773,
      "step": 554
    },
    {
      "epoch": 0.06530180021178962,
      "grad_norm": 5.039313316345215,
      "learning_rate": 1e-06,
      "loss": 0.8538,
      "step": 555
    },
    {
      "epoch": 0.06541946111307212,
      "grad_norm": 4.504653453826904,
      "learning_rate": 1e-06,
      "loss": 0.8726,
      "step": 556
    },
    {
      "epoch": 0.06553712201435463,
      "grad_norm": 3.786172866821289,
      "learning_rate": 1e-06,
      "loss": 0.7502,
      "step": 557
    },
    {
      "epoch": 0.06565478291563713,
      "grad_norm": 4.442392349243164,
      "learning_rate": 1e-06,
      "loss": 0.8309,
      "step": 558
    },
    {
      "epoch": 0.06577244381691964,
      "grad_norm": 4.390598773956299,
      "learning_rate": 1e-06,
      "loss": 0.8857,
      "step": 559
    },
    {
      "epoch": 0.06589010471820214,
      "grad_norm": 2.5926570892333984,
      "learning_rate": 1e-06,
      "loss": 0.7914,
      "step": 560
    },
    {
      "epoch": 0.06600776561948464,
      "grad_norm": 8.310322761535645,
      "learning_rate": 1e-06,
      "loss": 0.7625,
      "step": 561
    },
    {
      "epoch": 0.06612542652076715,
      "grad_norm": 12.237494468688965,
      "learning_rate": 1e-06,
      "loss": 0.7558,
      "step": 562
    },
    {
      "epoch": 0.06624308742204965,
      "grad_norm": 5.965966701507568,
      "learning_rate": 1e-06,
      "loss": 0.656,
      "step": 563
    },
    {
      "epoch": 0.06636074832333215,
      "grad_norm": 6.457314491271973,
      "learning_rate": 1e-06,
      "loss": 0.8754,
      "step": 564
    },
    {
      "epoch": 0.06647840922461466,
      "grad_norm": 2.9990360736846924,
      "learning_rate": 1e-06,
      "loss": 0.6746,
      "step": 565
    },
    {
      "epoch": 0.06659607012589716,
      "grad_norm": 6.220990180969238,
      "learning_rate": 1e-06,
      "loss": 0.7538,
      "step": 566
    },
    {
      "epoch": 0.06671373102717967,
      "grad_norm": 13.471657752990723,
      "learning_rate": 1e-06,
      "loss": 0.8042,
      "step": 567
    },
    {
      "epoch": 0.06683139192846217,
      "grad_norm": 6.996799945831299,
      "learning_rate": 1e-06,
      "loss": 0.8182,
      "step": 568
    },
    {
      "epoch": 0.06694905282974467,
      "grad_norm": 7.256328105926514,
      "learning_rate": 1e-06,
      "loss": 0.8296,
      "step": 569
    },
    {
      "epoch": 0.06706671373102718,
      "grad_norm": 6.098526477813721,
      "learning_rate": 1e-06,
      "loss": 0.7973,
      "step": 570
    },
    {
      "epoch": 0.06718437463230968,
      "grad_norm": 9.667437553405762,
      "learning_rate": 1e-06,
      "loss": 0.7141,
      "step": 571
    },
    {
      "epoch": 0.06730203553359218,
      "grad_norm": 4.498554229736328,
      "learning_rate": 1e-06,
      "loss": 0.9565,
      "step": 572
    },
    {
      "epoch": 0.06741969643487469,
      "grad_norm": 5.443784713745117,
      "learning_rate": 1e-06,
      "loss": 0.8814,
      "step": 573
    },
    {
      "epoch": 0.06753735733615719,
      "grad_norm": 4.860910415649414,
      "learning_rate": 1e-06,
      "loss": 0.6508,
      "step": 574
    },
    {
      "epoch": 0.0676550182374397,
      "grad_norm": 4.808187961578369,
      "learning_rate": 1e-06,
      "loss": 0.8269,
      "step": 575
    },
    {
      "epoch": 0.0677726791387222,
      "grad_norm": 4.875802993774414,
      "learning_rate": 1e-06,
      "loss": 0.7894,
      "step": 576
    },
    {
      "epoch": 0.0678903400400047,
      "grad_norm": 4.116751670837402,
      "learning_rate": 1e-06,
      "loss": 0.8942,
      "step": 577
    },
    {
      "epoch": 0.06800800094128721,
      "grad_norm": 3.51214599609375,
      "learning_rate": 1e-06,
      "loss": 0.6998,
      "step": 578
    },
    {
      "epoch": 0.06812566184256971,
      "grad_norm": 4.270370006561279,
      "learning_rate": 1e-06,
      "loss": 0.8708,
      "step": 579
    },
    {
      "epoch": 0.06824332274385221,
      "grad_norm": 3.15966796875,
      "learning_rate": 1e-06,
      "loss": 0.7595,
      "step": 580
    },
    {
      "epoch": 0.06836098364513472,
      "grad_norm": 4.910851955413818,
      "learning_rate": 1e-06,
      "loss": 0.6712,
      "step": 581
    },
    {
      "epoch": 0.06847864454641722,
      "grad_norm": 6.7051849365234375,
      "learning_rate": 1e-06,
      "loss": 0.8669,
      "step": 582
    },
    {
      "epoch": 0.06859630544769973,
      "grad_norm": 4.1636552810668945,
      "learning_rate": 1e-06,
      "loss": 0.8633,
      "step": 583
    },
    {
      "epoch": 0.06871396634898223,
      "grad_norm": 3.5476083755493164,
      "learning_rate": 1e-06,
      "loss": 0.8042,
      "step": 584
    },
    {
      "epoch": 0.06883162725026473,
      "grad_norm": 2.891711711883545,
      "learning_rate": 1e-06,
      "loss": 0.965,
      "step": 585
    },
    {
      "epoch": 0.06894928815154724,
      "grad_norm": 9.514786720275879,
      "learning_rate": 1e-06,
      "loss": 0.8521,
      "step": 586
    },
    {
      "epoch": 0.06906694905282974,
      "grad_norm": 4.414220809936523,
      "learning_rate": 1e-06,
      "loss": 0.8345,
      "step": 587
    },
    {
      "epoch": 0.06918460995411224,
      "grad_norm": 3.0958492755889893,
      "learning_rate": 1e-06,
      "loss": 0.7658,
      "step": 588
    },
    {
      "epoch": 0.06930227085539475,
      "grad_norm": 3.3460915088653564,
      "learning_rate": 1e-06,
      "loss": 0.8396,
      "step": 589
    },
    {
      "epoch": 0.06941993175667725,
      "grad_norm": 5.934871196746826,
      "learning_rate": 1e-06,
      "loss": 0.7246,
      "step": 590
    },
    {
      "epoch": 0.06953759265795977,
      "grad_norm": 5.69331169128418,
      "learning_rate": 1e-06,
      "loss": 0.7282,
      "step": 591
    },
    {
      "epoch": 0.06965525355924226,
      "grad_norm": 4.256507396697998,
      "learning_rate": 1e-06,
      "loss": 0.7915,
      "step": 592
    },
    {
      "epoch": 0.06977291446052476,
      "grad_norm": 5.126326084136963,
      "learning_rate": 1e-06,
      "loss": 0.7192,
      "step": 593
    },
    {
      "epoch": 0.06989057536180728,
      "grad_norm": 4.735490322113037,
      "learning_rate": 1e-06,
      "loss": 0.7529,
      "step": 594
    },
    {
      "epoch": 0.07000823626308977,
      "grad_norm": 3.5951437950134277,
      "learning_rate": 1e-06,
      "loss": 0.8213,
      "step": 595
    },
    {
      "epoch": 0.07012589716437227,
      "grad_norm": 6.8590168952941895,
      "learning_rate": 1e-06,
      "loss": 0.7228,
      "step": 596
    },
    {
      "epoch": 0.07024355806565478,
      "grad_norm": 5.810555934906006,
      "learning_rate": 1e-06,
      "loss": 0.747,
      "step": 597
    },
    {
      "epoch": 0.07036121896693728,
      "grad_norm": 4.012254238128662,
      "learning_rate": 1e-06,
      "loss": 0.7881,
      "step": 598
    },
    {
      "epoch": 0.0704788798682198,
      "grad_norm": 6.70487117767334,
      "learning_rate": 1e-06,
      "loss": 0.8001,
      "step": 599
    },
    {
      "epoch": 0.0705965407695023,
      "grad_norm": 2.8604013919830322,
      "learning_rate": 1e-06,
      "loss": 0.8036,
      "step": 600
    },
    {
      "epoch": 0.0707142016707848,
      "grad_norm": 3.871709108352661,
      "learning_rate": 1e-06,
      "loss": 0.6838,
      "step": 601
    },
    {
      "epoch": 0.0708318625720673,
      "grad_norm": 3.1215546131134033,
      "learning_rate": 1e-06,
      "loss": 0.802,
      "step": 602
    },
    {
      "epoch": 0.0709495234733498,
      "grad_norm": 2.95503306388855,
      "learning_rate": 1e-06,
      "loss": 0.7952,
      "step": 603
    },
    {
      "epoch": 0.0710671843746323,
      "grad_norm": 5.121756076812744,
      "learning_rate": 1e-06,
      "loss": 0.7197,
      "step": 604
    },
    {
      "epoch": 0.07118484527591482,
      "grad_norm": 3.5067391395568848,
      "learning_rate": 1e-06,
      "loss": 0.7953,
      "step": 605
    },
    {
      "epoch": 0.07130250617719731,
      "grad_norm": 2.0713229179382324,
      "learning_rate": 1e-06,
      "loss": 0.7853,
      "step": 606
    },
    {
      "epoch": 0.07142016707847983,
      "grad_norm": 3.8708510398864746,
      "learning_rate": 1e-06,
      "loss": 0.809,
      "step": 607
    },
    {
      "epoch": 0.07153782797976233,
      "grad_norm": 3.114241123199463,
      "learning_rate": 1e-06,
      "loss": 0.7183,
      "step": 608
    },
    {
      "epoch": 0.07165548888104482,
      "grad_norm": 2.5814106464385986,
      "learning_rate": 1e-06,
      "loss": 0.7672,
      "step": 609
    },
    {
      "epoch": 0.07177314978232734,
      "grad_norm": 5.265275001525879,
      "learning_rate": 1e-06,
      "loss": 0.7487,
      "step": 610
    },
    {
      "epoch": 0.07189081068360983,
      "grad_norm": 5.663961410522461,
      "learning_rate": 1e-06,
      "loss": 0.7835,
      "step": 611
    },
    {
      "epoch": 0.07200847158489233,
      "grad_norm": 11.934410095214844,
      "learning_rate": 1e-06,
      "loss": 0.8255,
      "step": 612
    },
    {
      "epoch": 0.07212613248617485,
      "grad_norm": 5.278149604797363,
      "learning_rate": 1e-06,
      "loss": 0.8952,
      "step": 613
    },
    {
      "epoch": 0.07224379338745734,
      "grad_norm": 5.30377721786499,
      "learning_rate": 1e-06,
      "loss": 0.8394,
      "step": 614
    },
    {
      "epoch": 0.07236145428873986,
      "grad_norm": 3.567551851272583,
      "learning_rate": 1e-06,
      "loss": 0.8797,
      "step": 615
    },
    {
      "epoch": 0.07247911519002236,
      "grad_norm": 5.296813011169434,
      "learning_rate": 1e-06,
      "loss": 0.8361,
      "step": 616
    },
    {
      "epoch": 0.07259677609130485,
      "grad_norm": 4.579360008239746,
      "learning_rate": 1e-06,
      "loss": 0.7945,
      "step": 617
    },
    {
      "epoch": 0.07271443699258737,
      "grad_norm": 4.319943904876709,
      "learning_rate": 1e-06,
      "loss": 0.7434,
      "step": 618
    },
    {
      "epoch": 0.07283209789386987,
      "grad_norm": 6.2155914306640625,
      "learning_rate": 1e-06,
      "loss": 0.7999,
      "step": 619
    },
    {
      "epoch": 0.07294975879515236,
      "grad_norm": 4.133585453033447,
      "learning_rate": 1e-06,
      "loss": 0.8252,
      "step": 620
    },
    {
      "epoch": 0.07306741969643488,
      "grad_norm": 12.52299976348877,
      "learning_rate": 1e-06,
      "loss": 0.6713,
      "step": 621
    },
    {
      "epoch": 0.07318508059771738,
      "grad_norm": 3.280254602432251,
      "learning_rate": 1e-06,
      "loss": 0.8483,
      "step": 622
    },
    {
      "epoch": 0.07330274149899989,
      "grad_norm": 4.4010138511657715,
      "learning_rate": 1e-06,
      "loss": 0.8613,
      "step": 623
    },
    {
      "epoch": 0.07342040240028239,
      "grad_norm": 3.92676043510437,
      "learning_rate": 1e-06,
      "loss": 0.8764,
      "step": 624
    },
    {
      "epoch": 0.07353806330156488,
      "grad_norm": 6.646303176879883,
      "learning_rate": 1e-06,
      "loss": 0.6661,
      "step": 625
    },
    {
      "epoch": 0.0736557242028474,
      "grad_norm": 4.366452217102051,
      "learning_rate": 1e-06,
      "loss": 0.8682,
      "step": 626
    },
    {
      "epoch": 0.0737733851041299,
      "grad_norm": 4.757740497589111,
      "learning_rate": 1e-06,
      "loss": 0.7562,
      "step": 627
    },
    {
      "epoch": 0.0738910460054124,
      "grad_norm": 4.1491241455078125,
      "learning_rate": 1e-06,
      "loss": 0.7249,
      "step": 628
    },
    {
      "epoch": 0.07400870690669491,
      "grad_norm": 3.2377617359161377,
      "learning_rate": 1e-06,
      "loss": 0.8469,
      "step": 629
    },
    {
      "epoch": 0.0741263678079774,
      "grad_norm": 3.7027909755706787,
      "learning_rate": 1e-06,
      "loss": 0.7918,
      "step": 630
    },
    {
      "epoch": 0.07424402870925992,
      "grad_norm": 4.722092628479004,
      "learning_rate": 1e-06,
      "loss": 0.8434,
      "step": 631
    },
    {
      "epoch": 0.07436168961054242,
      "grad_norm": 4.72484827041626,
      "learning_rate": 1e-06,
      "loss": 0.8254,
      "step": 632
    },
    {
      "epoch": 0.07447935051182492,
      "grad_norm": 4.173333644866943,
      "learning_rate": 1e-06,
      "loss": 0.7486,
      "step": 633
    },
    {
      "epoch": 0.07459701141310743,
      "grad_norm": 3.259687900543213,
      "learning_rate": 1e-06,
      "loss": 0.804,
      "step": 634
    },
    {
      "epoch": 0.07471467231438993,
      "grad_norm": 4.720432281494141,
      "learning_rate": 1e-06,
      "loss": 0.8884,
      "step": 635
    },
    {
      "epoch": 0.07483233321567243,
      "grad_norm": 2.8184030055999756,
      "learning_rate": 1e-06,
      "loss": 0.8045,
      "step": 636
    },
    {
      "epoch": 0.07494999411695494,
      "grad_norm": 2.802110433578491,
      "learning_rate": 1e-06,
      "loss": 0.8666,
      "step": 637
    },
    {
      "epoch": 0.07506765501823744,
      "grad_norm": 2.865649461746216,
      "learning_rate": 1e-06,
      "loss": 0.6902,
      "step": 638
    },
    {
      "epoch": 0.07518531591951995,
      "grad_norm": 3.1287262439727783,
      "learning_rate": 1e-06,
      "loss": 0.8662,
      "step": 639
    },
    {
      "epoch": 0.07530297682080245,
      "grad_norm": 5.535600662231445,
      "learning_rate": 1e-06,
      "loss": 0.7908,
      "step": 640
    },
    {
      "epoch": 0.07542063772208495,
      "grad_norm": 3.6541616916656494,
      "learning_rate": 1e-06,
      "loss": 0.843,
      "step": 641
    },
    {
      "epoch": 0.07553829862336746,
      "grad_norm": 3.577301502227783,
      "learning_rate": 1e-06,
      "loss": 0.7362,
      "step": 642
    },
    {
      "epoch": 0.07565595952464996,
      "grad_norm": 5.592081546783447,
      "learning_rate": 1e-06,
      "loss": 0.8034,
      "step": 643
    },
    {
      "epoch": 0.07577362042593246,
      "grad_norm": 3.633941173553467,
      "learning_rate": 1e-06,
      "loss": 0.8286,
      "step": 644
    },
    {
      "epoch": 0.07589128132721497,
      "grad_norm": 3.2601206302642822,
      "learning_rate": 1e-06,
      "loss": 0.7762,
      "step": 645
    },
    {
      "epoch": 0.07600894222849747,
      "grad_norm": 4.561715602874756,
      "learning_rate": 1e-06,
      "loss": 0.8529,
      "step": 646
    },
    {
      "epoch": 0.07612660312977998,
      "grad_norm": 9.033065795898438,
      "learning_rate": 1e-06,
      "loss": 0.8811,
      "step": 647
    },
    {
      "epoch": 0.07624426403106248,
      "grad_norm": 3.58722186088562,
      "learning_rate": 1e-06,
      "loss": 0.7805,
      "step": 648
    },
    {
      "epoch": 0.07636192493234498,
      "grad_norm": 2.476853370666504,
      "learning_rate": 1e-06,
      "loss": 0.7164,
      "step": 649
    },
    {
      "epoch": 0.07647958583362749,
      "grad_norm": 4.5690693855285645,
      "learning_rate": 1e-06,
      "loss": 0.6962,
      "step": 650
    },
    {
      "epoch": 0.07659724673490999,
      "grad_norm": 3.0442702770233154,
      "learning_rate": 1e-06,
      "loss": 0.7169,
      "step": 651
    },
    {
      "epoch": 0.07671490763619249,
      "grad_norm": 4.649383544921875,
      "learning_rate": 1e-06,
      "loss": 0.8122,
      "step": 652
    },
    {
      "epoch": 0.076832568537475,
      "grad_norm": 4.151635646820068,
      "learning_rate": 1e-06,
      "loss": 0.7678,
      "step": 653
    },
    {
      "epoch": 0.0769502294387575,
      "grad_norm": 5.340624809265137,
      "learning_rate": 1e-06,
      "loss": 0.7698,
      "step": 654
    },
    {
      "epoch": 0.07706789034004001,
      "grad_norm": 4.39862585067749,
      "learning_rate": 1e-06,
      "loss": 0.8883,
      "step": 655
    },
    {
      "epoch": 0.07718555124132251,
      "grad_norm": 7.796479225158691,
      "learning_rate": 1e-06,
      "loss": 0.7936,
      "step": 656
    },
    {
      "epoch": 0.07730321214260501,
      "grad_norm": 3.842503786087036,
      "learning_rate": 1e-06,
      "loss": 0.8487,
      "step": 657
    },
    {
      "epoch": 0.07742087304388752,
      "grad_norm": 12.752018928527832,
      "learning_rate": 1e-06,
      "loss": 0.8161,
      "step": 658
    },
    {
      "epoch": 0.07753853394517002,
      "grad_norm": 2.713529586791992,
      "learning_rate": 1e-06,
      "loss": 0.9158,
      "step": 659
    },
    {
      "epoch": 0.07765619484645252,
      "grad_norm": 5.894014358520508,
      "learning_rate": 1e-06,
      "loss": 0.927,
      "step": 660
    },
    {
      "epoch": 0.07777385574773503,
      "grad_norm": 10.467254638671875,
      "learning_rate": 1e-06,
      "loss": 0.8401,
      "step": 661
    },
    {
      "epoch": 0.07789151664901753,
      "grad_norm": 2.005702257156372,
      "learning_rate": 1e-06,
      "loss": 0.9121,
      "step": 662
    },
    {
      "epoch": 0.07800917755030004,
      "grad_norm": 2.868652582168579,
      "learning_rate": 1e-06,
      "loss": 0.7022,
      "step": 663
    },
    {
      "epoch": 0.07812683845158254,
      "grad_norm": 4.538110256195068,
      "learning_rate": 1e-06,
      "loss": 0.7856,
      "step": 664
    },
    {
      "epoch": 0.07824449935286504,
      "grad_norm": 3.814251184463501,
      "learning_rate": 1e-06,
      "loss": 0.7137,
      "step": 665
    },
    {
      "epoch": 0.07836216025414755,
      "grad_norm": 3.6797707080841064,
      "learning_rate": 1e-06,
      "loss": 0.6952,
      "step": 666
    },
    {
      "epoch": 0.07847982115543005,
      "grad_norm": 4.604442119598389,
      "learning_rate": 1e-06,
      "loss": 0.6523,
      "step": 667
    },
    {
      "epoch": 0.07859748205671255,
      "grad_norm": 6.55330753326416,
      "learning_rate": 1e-06,
      "loss": 0.6899,
      "step": 668
    },
    {
      "epoch": 0.07871514295799506,
      "grad_norm": 9.301312446594238,
      "learning_rate": 1e-06,
      "loss": 0.8195,
      "step": 669
    },
    {
      "epoch": 0.07883280385927756,
      "grad_norm": 4.445116996765137,
      "learning_rate": 1e-06,
      "loss": 0.7644,
      "step": 670
    },
    {
      "epoch": 0.07895046476056007,
      "grad_norm": 3.1115548610687256,
      "learning_rate": 1e-06,
      "loss": 0.788,
      "step": 671
    },
    {
      "epoch": 0.07906812566184257,
      "grad_norm": 3.480494499206543,
      "learning_rate": 1e-06,
      "loss": 0.8327,
      "step": 672
    },
    {
      "epoch": 0.07918578656312507,
      "grad_norm": 3.3543670177459717,
      "learning_rate": 1e-06,
      "loss": 0.9078,
      "step": 673
    },
    {
      "epoch": 0.07930344746440758,
      "grad_norm": 3.7252585887908936,
      "learning_rate": 1e-06,
      "loss": 0.6805,
      "step": 674
    },
    {
      "epoch": 0.07942110836569008,
      "grad_norm": 3.1235790252685547,
      "learning_rate": 1e-06,
      "loss": 0.7886,
      "step": 675
    },
    {
      "epoch": 0.07953876926697258,
      "grad_norm": 2.5716028213500977,
      "learning_rate": 1e-06,
      "loss": 0.7839,
      "step": 676
    },
    {
      "epoch": 0.07965643016825509,
      "grad_norm": 5.055776119232178,
      "learning_rate": 1e-06,
      "loss": 0.7922,
      "step": 677
    },
    {
      "epoch": 0.07977409106953759,
      "grad_norm": 6.259873867034912,
      "learning_rate": 1e-06,
      "loss": 0.6882,
      "step": 678
    },
    {
      "epoch": 0.0798917519708201,
      "grad_norm": 6.074028968811035,
      "learning_rate": 1e-06,
      "loss": 0.8171,
      "step": 679
    },
    {
      "epoch": 0.0800094128721026,
      "grad_norm": 6.025778293609619,
      "learning_rate": 1e-06,
      "loss": 0.7963,
      "step": 680
    },
    {
      "epoch": 0.0801270737733851,
      "grad_norm": 4.927762508392334,
      "learning_rate": 1e-06,
      "loss": 0.7133,
      "step": 681
    },
    {
      "epoch": 0.08024473467466761,
      "grad_norm": 9.872488021850586,
      "learning_rate": 1e-06,
      "loss": 0.8878,
      "step": 682
    },
    {
      "epoch": 0.08036239557595011,
      "grad_norm": 4.145512580871582,
      "learning_rate": 1e-06,
      "loss": 0.6282,
      "step": 683
    },
    {
      "epoch": 0.08048005647723261,
      "grad_norm": 4.067215442657471,
      "learning_rate": 1e-06,
      "loss": 0.8972,
      "step": 684
    },
    {
      "epoch": 0.08059771737851512,
      "grad_norm": 6.702483177185059,
      "learning_rate": 1e-06,
      "loss": 0.71,
      "step": 685
    },
    {
      "epoch": 0.08071537827979762,
      "grad_norm": 7.621339797973633,
      "learning_rate": 1e-06,
      "loss": 0.7837,
      "step": 686
    },
    {
      "epoch": 0.08083303918108013,
      "grad_norm": 2.6410186290740967,
      "learning_rate": 1e-06,
      "loss": 0.7883,
      "step": 687
    },
    {
      "epoch": 0.08095070008236263,
      "grad_norm": 5.255964756011963,
      "learning_rate": 1e-06,
      "loss": 0.8354,
      "step": 688
    },
    {
      "epoch": 0.08106836098364513,
      "grad_norm": 4.244592666625977,
      "learning_rate": 1e-06,
      "loss": 0.8109,
      "step": 689
    },
    {
      "epoch": 0.08118602188492764,
      "grad_norm": 2.6107852458953857,
      "learning_rate": 1e-06,
      "loss": 0.8591,
      "step": 690
    },
    {
      "epoch": 0.08130368278621014,
      "grad_norm": 4.650479316711426,
      "learning_rate": 1e-06,
      "loss": 0.6969,
      "step": 691
    },
    {
      "epoch": 0.08142134368749264,
      "grad_norm": 3.1627695560455322,
      "learning_rate": 1e-06,
      "loss": 0.8289,
      "step": 692
    },
    {
      "epoch": 0.08153900458877515,
      "grad_norm": 4.6385674476623535,
      "learning_rate": 1e-06,
      "loss": 0.7032,
      "step": 693
    },
    {
      "epoch": 0.08165666549005765,
      "grad_norm": 5.22135066986084,
      "learning_rate": 1e-06,
      "loss": 0.7547,
      "step": 694
    },
    {
      "epoch": 0.08177432639134016,
      "grad_norm": 4.280886173248291,
      "learning_rate": 1e-06,
      "loss": 0.8054,
      "step": 695
    },
    {
      "epoch": 0.08189198729262266,
      "grad_norm": 10.859444618225098,
      "learning_rate": 1e-06,
      "loss": 0.7395,
      "step": 696
    },
    {
      "epoch": 0.08200964819390516,
      "grad_norm": 3.9999287128448486,
      "learning_rate": 1e-06,
      "loss": 0.7837,
      "step": 697
    },
    {
      "epoch": 0.08212730909518767,
      "grad_norm": 7.206291198730469,
      "learning_rate": 1e-06,
      "loss": 0.764,
      "step": 698
    },
    {
      "epoch": 0.08224496999647017,
      "grad_norm": 5.005997180938721,
      "learning_rate": 1e-06,
      "loss": 0.923,
      "step": 699
    },
    {
      "epoch": 0.08236263089775267,
      "grad_norm": 4.134511470794678,
      "learning_rate": 1e-06,
      "loss": 0.815,
      "step": 700
    },
    {
      "epoch": 0.08248029179903518,
      "grad_norm": 4.48399543762207,
      "learning_rate": 1e-06,
      "loss": 0.8545,
      "step": 701
    },
    {
      "epoch": 0.08259795270031768,
      "grad_norm": 7.961074352264404,
      "learning_rate": 1e-06,
      "loss": 0.8463,
      "step": 702
    },
    {
      "epoch": 0.0827156136016002,
      "grad_norm": 3.7703304290771484,
      "learning_rate": 1e-06,
      "loss": 0.7455,
      "step": 703
    },
    {
      "epoch": 0.08283327450288269,
      "grad_norm": 4.462988376617432,
      "learning_rate": 1e-06,
      "loss": 0.8649,
      "step": 704
    },
    {
      "epoch": 0.08295093540416519,
      "grad_norm": 5.397121429443359,
      "learning_rate": 1e-06,
      "loss": 0.8086,
      "step": 705
    },
    {
      "epoch": 0.0830685963054477,
      "grad_norm": 3.6075279712677,
      "learning_rate": 1e-06,
      "loss": 0.7628,
      "step": 706
    },
    {
      "epoch": 0.0831862572067302,
      "grad_norm": 3.9430606365203857,
      "learning_rate": 1e-06,
      "loss": 0.8156,
      "step": 707
    },
    {
      "epoch": 0.0833039181080127,
      "grad_norm": 3.9803953170776367,
      "learning_rate": 1e-06,
      "loss": 0.7861,
      "step": 708
    },
    {
      "epoch": 0.08342157900929521,
      "grad_norm": 5.1148762702941895,
      "learning_rate": 1e-06,
      "loss": 0.8435,
      "step": 709
    },
    {
      "epoch": 0.08353923991057771,
      "grad_norm": 3.512322425842285,
      "learning_rate": 1e-06,
      "loss": 0.8721,
      "step": 710
    },
    {
      "epoch": 0.08365690081186022,
      "grad_norm": 3.4553308486938477,
      "learning_rate": 1e-06,
      "loss": 0.8139,
      "step": 711
    },
    {
      "epoch": 0.08377456171314272,
      "grad_norm": 5.630927085876465,
      "learning_rate": 1e-06,
      "loss": 0.8191,
      "step": 712
    },
    {
      "epoch": 0.08389222261442522,
      "grad_norm": 10.072120666503906,
      "learning_rate": 1e-06,
      "loss": 0.8147,
      "step": 713
    },
    {
      "epoch": 0.08400988351570773,
      "grad_norm": 3.564073324203491,
      "learning_rate": 1e-06,
      "loss": 0.8458,
      "step": 714
    },
    {
      "epoch": 0.08412754441699023,
      "grad_norm": 9.791050910949707,
      "learning_rate": 1e-06,
      "loss": 0.9003,
      "step": 715
    },
    {
      "epoch": 0.08424520531827273,
      "grad_norm": 11.41666316986084,
      "learning_rate": 1e-06,
      "loss": 0.8637,
      "step": 716
    },
    {
      "epoch": 0.08436286621955524,
      "grad_norm": 9.25336742401123,
      "learning_rate": 1e-06,
      "loss": 0.8096,
      "step": 717
    },
    {
      "epoch": 0.08448052712083774,
      "grad_norm": 3.4500648975372314,
      "learning_rate": 1e-06,
      "loss": 0.7901,
      "step": 718
    },
    {
      "epoch": 0.08459818802212025,
      "grad_norm": 3.696110725402832,
      "learning_rate": 1e-06,
      "loss": 0.6153,
      "step": 719
    },
    {
      "epoch": 0.08471584892340275,
      "grad_norm": 5.816553115844727,
      "learning_rate": 1e-06,
      "loss": 0.9023,
      "step": 720
    },
    {
      "epoch": 0.08483350982468525,
      "grad_norm": 5.944538116455078,
      "learning_rate": 1e-06,
      "loss": 0.7745,
      "step": 721
    },
    {
      "epoch": 0.08495117072596776,
      "grad_norm": 9.655933380126953,
      "learning_rate": 1e-06,
      "loss": 0.7756,
      "step": 722
    },
    {
      "epoch": 0.08506883162725026,
      "grad_norm": 4.010641574859619,
      "learning_rate": 1e-06,
      "loss": 0.7837,
      "step": 723
    },
    {
      "epoch": 0.08518649252853276,
      "grad_norm": 3.5823662281036377,
      "learning_rate": 1e-06,
      "loss": 0.7036,
      "step": 724
    },
    {
      "epoch": 0.08530415342981527,
      "grad_norm": 3.164689302444458,
      "learning_rate": 1e-06,
      "loss": 0.8682,
      "step": 725
    },
    {
      "epoch": 0.08542181433109777,
      "grad_norm": 3.7373437881469727,
      "learning_rate": 1e-06,
      "loss": 0.8604,
      "step": 726
    },
    {
      "epoch": 0.08553947523238029,
      "grad_norm": 3.8801088333129883,
      "learning_rate": 1e-06,
      "loss": 0.7022,
      "step": 727
    },
    {
      "epoch": 0.08565713613366278,
      "grad_norm": 3.8801088333129883,
      "learning_rate": 1e-06,
      "loss": 0.8586,
      "step": 728
    },
    {
      "epoch": 0.08577479703494528,
      "grad_norm": 4.299330234527588,
      "learning_rate": 1e-06,
      "loss": 0.7284,
      "step": 729
    },
    {
      "epoch": 0.0858924579362278,
      "grad_norm": 12.118149757385254,
      "learning_rate": 1e-06,
      "loss": 0.7516,
      "step": 730
    },
    {
      "epoch": 0.0860101188375103,
      "grad_norm": 4.249277114868164,
      "learning_rate": 1e-06,
      "loss": 0.804,
      "step": 731
    },
    {
      "epoch": 0.08612777973879279,
      "grad_norm": 3.9434337615966797,
      "learning_rate": 1e-06,
      "loss": 0.7754,
      "step": 732
    },
    {
      "epoch": 0.0862454406400753,
      "grad_norm": 9.427238464355469,
      "learning_rate": 1e-06,
      "loss": 0.7556,
      "step": 733
    },
    {
      "epoch": 0.0863631015413578,
      "grad_norm": 7.601968288421631,
      "learning_rate": 1e-06,
      "loss": 0.9026,
      "step": 734
    },
    {
      "epoch": 0.08648076244264032,
      "grad_norm": 4.908003807067871,
      "learning_rate": 1e-06,
      "loss": 0.7717,
      "step": 735
    },
    {
      "epoch": 0.08659842334392281,
      "grad_norm": 5.9103522300720215,
      "learning_rate": 1e-06,
      "loss": 0.8005,
      "step": 736
    },
    {
      "epoch": 0.08671608424520531,
      "grad_norm": 3.8755075931549072,
      "learning_rate": 1e-06,
      "loss": 0.7784,
      "step": 737
    },
    {
      "epoch": 0.08683374514648783,
      "grad_norm": 5.542903423309326,
      "learning_rate": 1e-06,
      "loss": 0.7552,
      "step": 738
    },
    {
      "epoch": 0.08695140604777032,
      "grad_norm": 3.3967931270599365,
      "learning_rate": 1e-06,
      "loss": 0.738,
      "step": 739
    },
    {
      "epoch": 0.08706906694905282,
      "grad_norm": 2.3593356609344482,
      "learning_rate": 1e-06,
      "loss": 0.7664,
      "step": 740
    },
    {
      "epoch": 0.08718672785033534,
      "grad_norm": 3.30541729927063,
      "learning_rate": 1e-06,
      "loss": 0.798,
      "step": 741
    },
    {
      "epoch": 0.08730438875161783,
      "grad_norm": 7.929763317108154,
      "learning_rate": 1e-06,
      "loss": 0.7142,
      "step": 742
    },
    {
      "epoch": 0.08742204965290035,
      "grad_norm": 4.586026191711426,
      "learning_rate": 1e-06,
      "loss": 0.8118,
      "step": 743
    },
    {
      "epoch": 0.08753971055418285,
      "grad_norm": 3.0885305404663086,
      "learning_rate": 1e-06,
      "loss": 0.831,
      "step": 744
    },
    {
      "epoch": 0.08765737145546534,
      "grad_norm": 6.417020797729492,
      "learning_rate": 1e-06,
      "loss": 0.7054,
      "step": 745
    },
    {
      "epoch": 0.08777503235674786,
      "grad_norm": 8.70758056640625,
      "learning_rate": 1e-06,
      "loss": 0.8487,
      "step": 746
    },
    {
      "epoch": 0.08789269325803035,
      "grad_norm": 5.649046421051025,
      "learning_rate": 1e-06,
      "loss": 0.8484,
      "step": 747
    },
    {
      "epoch": 0.08801035415931285,
      "grad_norm": 2.783040761947632,
      "learning_rate": 1e-06,
      "loss": 0.731,
      "step": 748
    },
    {
      "epoch": 0.08812801506059537,
      "grad_norm": 3.5551412105560303,
      "learning_rate": 1e-06,
      "loss": 0.7613,
      "step": 749
    },
    {
      "epoch": 0.08824567596187786,
      "grad_norm": 3.094209671020508,
      "learning_rate": 1e-06,
      "loss": 0.8581,
      "step": 750
    },
    {
      "epoch": 0.08836333686316038,
      "grad_norm": 6.274815082550049,
      "learning_rate": 1e-06,
      "loss": 0.8412,
      "step": 751
    },
    {
      "epoch": 0.08848099776444288,
      "grad_norm": 5.962408542633057,
      "learning_rate": 1e-06,
      "loss": 0.8323,
      "step": 752
    },
    {
      "epoch": 0.08859865866572537,
      "grad_norm": 2.927711009979248,
      "learning_rate": 1e-06,
      "loss": 0.8187,
      "step": 753
    },
    {
      "epoch": 0.08871631956700789,
      "grad_norm": 7.680797576904297,
      "learning_rate": 1e-06,
      "loss": 0.7775,
      "step": 754
    },
    {
      "epoch": 0.08883398046829039,
      "grad_norm": 45.736412048339844,
      "learning_rate": 1e-06,
      "loss": 0.5749,
      "step": 755
    },
    {
      "epoch": 0.08895164136957288,
      "grad_norm": 5.9662017822265625,
      "learning_rate": 1e-06,
      "loss": 0.7663,
      "step": 756
    },
    {
      "epoch": 0.0890693022708554,
      "grad_norm": 3.3588130474090576,
      "learning_rate": 1e-06,
      "loss": 0.7206,
      "step": 757
    },
    {
      "epoch": 0.0891869631721379,
      "grad_norm": 4.659145832061768,
      "learning_rate": 1e-06,
      "loss": 0.8381,
      "step": 758
    },
    {
      "epoch": 0.08930462407342041,
      "grad_norm": 5.114917755126953,
      "learning_rate": 1e-06,
      "loss": 0.6961,
      "step": 759
    },
    {
      "epoch": 0.0894222849747029,
      "grad_norm": 6.852463722229004,
      "learning_rate": 1e-06,
      "loss": 0.7245,
      "step": 760
    },
    {
      "epoch": 0.0895399458759854,
      "grad_norm": 3.8735191822052,
      "learning_rate": 1e-06,
      "loss": 0.8026,
      "step": 761
    },
    {
      "epoch": 0.08965760677726792,
      "grad_norm": 3.054180860519409,
      "learning_rate": 1e-06,
      "loss": 0.903,
      "step": 762
    },
    {
      "epoch": 0.08977526767855042,
      "grad_norm": 18.586585998535156,
      "learning_rate": 1e-06,
      "loss": 0.7696,
      "step": 763
    },
    {
      "epoch": 0.08989292857983291,
      "grad_norm": 2.874756097793579,
      "learning_rate": 1e-06,
      "loss": 0.7986,
      "step": 764
    },
    {
      "epoch": 0.09001058948111543,
      "grad_norm": 1.945826530456543,
      "learning_rate": 1e-06,
      "loss": 0.841,
      "step": 765
    },
    {
      "epoch": 0.09012825038239793,
      "grad_norm": 3.9630308151245117,
      "learning_rate": 1e-06,
      "loss": 0.7335,
      "step": 766
    },
    {
      "epoch": 0.09024591128368044,
      "grad_norm": 5.930178165435791,
      "learning_rate": 1e-06,
      "loss": 0.7273,
      "step": 767
    },
    {
      "epoch": 0.09036357218496294,
      "grad_norm": 4.515182971954346,
      "learning_rate": 1e-06,
      "loss": 0.842,
      "step": 768
    },
    {
      "epoch": 0.09048123308624544,
      "grad_norm": 4.042099952697754,
      "learning_rate": 1e-06,
      "loss": 0.6113,
      "step": 769
    },
    {
      "epoch": 0.09059889398752795,
      "grad_norm": 3.2117252349853516,
      "learning_rate": 1e-06,
      "loss": 0.8613,
      "step": 770
    },
    {
      "epoch": 0.09071655488881045,
      "grad_norm": 3.9129526615142822,
      "learning_rate": 1e-06,
      "loss": 0.6426,
      "step": 771
    },
    {
      "epoch": 0.09083421579009295,
      "grad_norm": 3.7040510177612305,
      "learning_rate": 1e-06,
      "loss": 0.7795,
      "step": 772
    },
    {
      "epoch": 0.09095187669137546,
      "grad_norm": 5.53035306930542,
      "learning_rate": 1e-06,
      "loss": 0.6404,
      "step": 773
    },
    {
      "epoch": 0.09106953759265796,
      "grad_norm": 3.193643569946289,
      "learning_rate": 1e-06,
      "loss": 0.7391,
      "step": 774
    },
    {
      "epoch": 0.09118719849394047,
      "grad_norm": 2.597047805786133,
      "learning_rate": 1e-06,
      "loss": 0.9017,
      "step": 775
    },
    {
      "epoch": 0.09130485939522297,
      "grad_norm": 3.4000656604766846,
      "learning_rate": 1e-06,
      "loss": 0.7492,
      "step": 776
    },
    {
      "epoch": 0.09142252029650547,
      "grad_norm": 2.4344582557678223,
      "learning_rate": 1e-06,
      "loss": 0.7923,
      "step": 777
    },
    {
      "epoch": 0.09154018119778798,
      "grad_norm": 3.083674192428589,
      "learning_rate": 1e-06,
      "loss": 0.9369,
      "step": 778
    },
    {
      "epoch": 0.09165784209907048,
      "grad_norm": 5.288464069366455,
      "learning_rate": 1e-06,
      "loss": 0.68,
      "step": 779
    },
    {
      "epoch": 0.09177550300035298,
      "grad_norm": 3.390939474105835,
      "learning_rate": 1e-06,
      "loss": 0.7158,
      "step": 780
    },
    {
      "epoch": 0.09189316390163549,
      "grad_norm": 3.3622586727142334,
      "learning_rate": 1e-06,
      "loss": 0.8059,
      "step": 781
    },
    {
      "epoch": 0.09201082480291799,
      "grad_norm": 3.347780227661133,
      "learning_rate": 1e-06,
      "loss": 0.7351,
      "step": 782
    },
    {
      "epoch": 0.0921284857042005,
      "grad_norm": 4.869822978973389,
      "learning_rate": 1e-06,
      "loss": 0.705,
      "step": 783
    },
    {
      "epoch": 0.092246146605483,
      "grad_norm": 8.056663513183594,
      "learning_rate": 1e-06,
      "loss": 0.7346,
      "step": 784
    },
    {
      "epoch": 0.0923638075067655,
      "grad_norm": 4.730690002441406,
      "learning_rate": 1e-06,
      "loss": 0.8079,
      "step": 785
    },
    {
      "epoch": 0.09248146840804801,
      "grad_norm": 3.544837713241577,
      "learning_rate": 1e-06,
      "loss": 0.7663,
      "step": 786
    },
    {
      "epoch": 0.09259912930933051,
      "grad_norm": 4.1507649421691895,
      "learning_rate": 1e-06,
      "loss": 0.7674,
      "step": 787
    },
    {
      "epoch": 0.092716790210613,
      "grad_norm": 7.242065906524658,
      "learning_rate": 1e-06,
      "loss": 0.8021,
      "step": 788
    },
    {
      "epoch": 0.09283445111189552,
      "grad_norm": 6.251343250274658,
      "learning_rate": 1e-06,
      "loss": 0.7952,
      "step": 789
    },
    {
      "epoch": 0.09295211201317802,
      "grad_norm": 4.684685230255127,
      "learning_rate": 1e-06,
      "loss": 0.8428,
      "step": 790
    },
    {
      "epoch": 0.09306977291446053,
      "grad_norm": 3.2917494773864746,
      "learning_rate": 1e-06,
      "loss": 0.8165,
      "step": 791
    },
    {
      "epoch": 0.09318743381574303,
      "grad_norm": 3.4737491607666016,
      "learning_rate": 1e-06,
      "loss": 0.7606,
      "step": 792
    },
    {
      "epoch": 0.09330509471702553,
      "grad_norm": 2.5431602001190186,
      "learning_rate": 1e-06,
      "loss": 0.8489,
      "step": 793
    },
    {
      "epoch": 0.09342275561830804,
      "grad_norm": 4.048852443695068,
      "learning_rate": 1e-06,
      "loss": 0.7458,
      "step": 794
    },
    {
      "epoch": 0.09354041651959054,
      "grad_norm": 7.465090751647949,
      "learning_rate": 1e-06,
      "loss": 0.8441,
      "step": 795
    },
    {
      "epoch": 0.09365807742087304,
      "grad_norm": 5.336065769195557,
      "learning_rate": 1e-06,
      "loss": 0.7008,
      "step": 796
    },
    {
      "epoch": 0.09377573832215555,
      "grad_norm": 3.435615301132202,
      "learning_rate": 1e-06,
      "loss": 0.795,
      "step": 797
    },
    {
      "epoch": 0.09389339922343805,
      "grad_norm": 7.03810977935791,
      "learning_rate": 1e-06,
      "loss": 0.8487,
      "step": 798
    },
    {
      "epoch": 0.09401106012472056,
      "grad_norm": 3.2493505477905273,
      "learning_rate": 1e-06,
      "loss": 0.7925,
      "step": 799
    },
    {
      "epoch": 0.09412872102600306,
      "grad_norm": 4.32244348526001,
      "learning_rate": 1e-06,
      "loss": 0.9033,
      "step": 800
    },
    {
      "epoch": 0.09424638192728556,
      "grad_norm": 6.156367301940918,
      "learning_rate": 1e-06,
      "loss": 0.8354,
      "step": 801
    },
    {
      "epoch": 0.09436404282856807,
      "grad_norm": 3.897897481918335,
      "learning_rate": 1e-06,
      "loss": 0.667,
      "step": 802
    },
    {
      "epoch": 0.09448170372985057,
      "grad_norm": 4.081802845001221,
      "learning_rate": 1e-06,
      "loss": 0.8003,
      "step": 803
    },
    {
      "epoch": 0.09459936463113307,
      "grad_norm": 2.2866339683532715,
      "learning_rate": 1e-06,
      "loss": 0.9098,
      "step": 804
    },
    {
      "epoch": 0.09471702553241558,
      "grad_norm": 10.533675193786621,
      "learning_rate": 1e-06,
      "loss": 0.6993,
      "step": 805
    },
    {
      "epoch": 0.09483468643369808,
      "grad_norm": 2.7982473373413086,
      "learning_rate": 1e-06,
      "loss": 0.7198,
      "step": 806
    },
    {
      "epoch": 0.09495234733498059,
      "grad_norm": 6.184414386749268,
      "learning_rate": 1e-06,
      "loss": 0.7414,
      "step": 807
    },
    {
      "epoch": 0.09507000823626309,
      "grad_norm": 7.496194839477539,
      "learning_rate": 1e-06,
      "loss": 0.6699,
      "step": 808
    },
    {
      "epoch": 0.09518766913754559,
      "grad_norm": 2.7355363368988037,
      "learning_rate": 1e-06,
      "loss": 0.7179,
      "step": 809
    },
    {
      "epoch": 0.0953053300388281,
      "grad_norm": 4.605178356170654,
      "learning_rate": 1e-06,
      "loss": 0.7431,
      "step": 810
    },
    {
      "epoch": 0.0954229909401106,
      "grad_norm": 3.33636212348938,
      "learning_rate": 1e-06,
      "loss": 0.6955,
      "step": 811
    },
    {
      "epoch": 0.0955406518413931,
      "grad_norm": 6.929253578186035,
      "learning_rate": 1e-06,
      "loss": 0.7663,
      "step": 812
    },
    {
      "epoch": 0.09565831274267561,
      "grad_norm": 4.302491664886475,
      "learning_rate": 1e-06,
      "loss": 0.9389,
      "step": 813
    },
    {
      "epoch": 0.09577597364395811,
      "grad_norm": 4.514020919799805,
      "learning_rate": 1e-06,
      "loss": 0.778,
      "step": 814
    },
    {
      "epoch": 0.09589363454524062,
      "grad_norm": 3.7888290882110596,
      "learning_rate": 1e-06,
      "loss": 0.7926,
      "step": 815
    },
    {
      "epoch": 0.09601129544652312,
      "grad_norm": 4.159764766693115,
      "learning_rate": 1e-06,
      "loss": 0.6922,
      "step": 816
    },
    {
      "epoch": 0.09612895634780562,
      "grad_norm": 2.528047561645508,
      "learning_rate": 1e-06,
      "loss": 0.8371,
      "step": 817
    },
    {
      "epoch": 0.09624661724908813,
      "grad_norm": 5.680185794830322,
      "learning_rate": 1e-06,
      "loss": 0.6564,
      "step": 818
    },
    {
      "epoch": 0.09636427815037063,
      "grad_norm": 5.111583709716797,
      "learning_rate": 1e-06,
      "loss": 0.827,
      "step": 819
    },
    {
      "epoch": 0.09648193905165313,
      "grad_norm": 3.2532787322998047,
      "learning_rate": 1e-06,
      "loss": 0.8875,
      "step": 820
    },
    {
      "epoch": 0.09659959995293564,
      "grad_norm": 6.838262557983398,
      "learning_rate": 1e-06,
      "loss": 0.7568,
      "step": 821
    },
    {
      "epoch": 0.09671726085421814,
      "grad_norm": 5.076169013977051,
      "learning_rate": 1e-06,
      "loss": 0.8047,
      "step": 822
    },
    {
      "epoch": 0.09683492175550065,
      "grad_norm": 3.4474267959594727,
      "learning_rate": 1e-06,
      "loss": 0.7991,
      "step": 823
    },
    {
      "epoch": 0.09695258265678315,
      "grad_norm": 2.7672460079193115,
      "learning_rate": 1e-06,
      "loss": 0.8094,
      "step": 824
    },
    {
      "epoch": 0.09707024355806565,
      "grad_norm": 3.9804646968841553,
      "learning_rate": 1e-06,
      "loss": 0.7259,
      "step": 825
    },
    {
      "epoch": 0.09718790445934816,
      "grad_norm": 3.030790090560913,
      "learning_rate": 1e-06,
      "loss": 0.7429,
      "step": 826
    },
    {
      "epoch": 0.09730556536063066,
      "grad_norm": 5.696221828460693,
      "learning_rate": 1e-06,
      "loss": 0.7522,
      "step": 827
    },
    {
      "epoch": 0.09742322626191316,
      "grad_norm": 2.9806602001190186,
      "learning_rate": 1e-06,
      "loss": 0.7438,
      "step": 828
    },
    {
      "epoch": 0.09754088716319567,
      "grad_norm": 6.773803234100342,
      "learning_rate": 1e-06,
      "loss": 0.7772,
      "step": 829
    },
    {
      "epoch": 0.09765854806447817,
      "grad_norm": 3.165433883666992,
      "learning_rate": 1e-06,
      "loss": 0.7221,
      "step": 830
    },
    {
      "epoch": 0.09777620896576068,
      "grad_norm": 3.9233906269073486,
      "learning_rate": 1e-06,
      "loss": 0.8598,
      "step": 831
    },
    {
      "epoch": 0.09789386986704318,
      "grad_norm": 2.9941489696502686,
      "learning_rate": 1e-06,
      "loss": 0.6664,
      "step": 832
    },
    {
      "epoch": 0.09801153076832568,
      "grad_norm": 3.3423893451690674,
      "learning_rate": 1e-06,
      "loss": 0.8662,
      "step": 833
    },
    {
      "epoch": 0.09812919166960819,
      "grad_norm": 7.260572910308838,
      "learning_rate": 1e-06,
      "loss": 0.6569,
      "step": 834
    },
    {
      "epoch": 0.09824685257089069,
      "grad_norm": 21.946184158325195,
      "learning_rate": 1e-06,
      "loss": 0.7579,
      "step": 835
    },
    {
      "epoch": 0.09836451347217319,
      "grad_norm": 4.030398845672607,
      "learning_rate": 1e-06,
      "loss": 0.8875,
      "step": 836
    },
    {
      "epoch": 0.0984821743734557,
      "grad_norm": 5.5473833084106445,
      "learning_rate": 1e-06,
      "loss": 0.6712,
      "step": 837
    },
    {
      "epoch": 0.0985998352747382,
      "grad_norm": 2.977285146713257,
      "learning_rate": 1e-06,
      "loss": 0.6802,
      "step": 838
    },
    {
      "epoch": 0.09871749617602071,
      "grad_norm": 4.305202960968018,
      "learning_rate": 1e-06,
      "loss": 0.8286,
      "step": 839
    },
    {
      "epoch": 0.09883515707730321,
      "grad_norm": 4.800453186035156,
      "learning_rate": 1e-06,
      "loss": 0.7722,
      "step": 840
    },
    {
      "epoch": 0.09895281797858571,
      "grad_norm": 4.110867023468018,
      "learning_rate": 1e-06,
      "loss": 0.736,
      "step": 841
    },
    {
      "epoch": 0.09907047887986822,
      "grad_norm": 4.379387855529785,
      "learning_rate": 1e-06,
      "loss": 0.7814,
      "step": 842
    },
    {
      "epoch": 0.09918813978115072,
      "grad_norm": 5.1933274269104,
      "learning_rate": 1e-06,
      "loss": 0.8989,
      "step": 843
    },
    {
      "epoch": 0.09930580068243322,
      "grad_norm": 4.651576995849609,
      "learning_rate": 1e-06,
      "loss": 0.8684,
      "step": 844
    },
    {
      "epoch": 0.09942346158371573,
      "grad_norm": 3.9711875915527344,
      "learning_rate": 1e-06,
      "loss": 0.8008,
      "step": 845
    },
    {
      "epoch": 0.09954112248499823,
      "grad_norm": 4.233101844787598,
      "learning_rate": 1e-06,
      "loss": 0.767,
      "step": 846
    },
    {
      "epoch": 0.09965878338628074,
      "grad_norm": 4.948994159698486,
      "learning_rate": 1e-06,
      "loss": 0.7025,
      "step": 847
    },
    {
      "epoch": 0.09977644428756324,
      "grad_norm": 4.045925140380859,
      "learning_rate": 1e-06,
      "loss": 0.7014,
      "step": 848
    },
    {
      "epoch": 0.09989410518884574,
      "grad_norm": 4.309968948364258,
      "learning_rate": 1e-06,
      "loss": 0.8363,
      "step": 849
    },
    {
      "epoch": 0.10001176609012825,
      "grad_norm": 4.463519096374512,
      "learning_rate": 1e-06,
      "loss": 0.7713,
      "step": 850
    },
    {
      "epoch": 0.10012942699141075,
      "grad_norm": 5.474331378936768,
      "learning_rate": 1e-06,
      "loss": 0.7899,
      "step": 851
    },
    {
      "epoch": 0.10024708789269325,
      "grad_norm": 8.145475387573242,
      "learning_rate": 1e-06,
      "loss": 0.7552,
      "step": 852
    },
    {
      "epoch": 0.10036474879397576,
      "grad_norm": 9.849764823913574,
      "learning_rate": 1e-06,
      "loss": 0.6796,
      "step": 853
    },
    {
      "epoch": 0.10048240969525826,
      "grad_norm": 3.580988883972168,
      "learning_rate": 1e-06,
      "loss": 0.818,
      "step": 854
    },
    {
      "epoch": 0.10060007059654077,
      "grad_norm": 2.9809317588806152,
      "learning_rate": 1e-06,
      "loss": 0.8343,
      "step": 855
    },
    {
      "epoch": 0.10071773149782327,
      "grad_norm": 4.253994464874268,
      "learning_rate": 1e-06,
      "loss": 0.7922,
      "step": 856
    },
    {
      "epoch": 0.10083539239910577,
      "grad_norm": 7.390686988830566,
      "learning_rate": 1e-06,
      "loss": 0.8706,
      "step": 857
    },
    {
      "epoch": 0.10095305330038828,
      "grad_norm": 6.373108386993408,
      "learning_rate": 1e-06,
      "loss": 0.6844,
      "step": 858
    },
    {
      "epoch": 0.10107071420167078,
      "grad_norm": 5.433058738708496,
      "learning_rate": 1e-06,
      "loss": 0.7881,
      "step": 859
    },
    {
      "epoch": 0.10118837510295328,
      "grad_norm": 5.169571399688721,
      "learning_rate": 1e-06,
      "loss": 0.7737,
      "step": 860
    },
    {
      "epoch": 0.1013060360042358,
      "grad_norm": 3.5316271781921387,
      "learning_rate": 1e-06,
      "loss": 0.7511,
      "step": 861
    },
    {
      "epoch": 0.10142369690551829,
      "grad_norm": 3.4881837368011475,
      "learning_rate": 1e-06,
      "loss": 0.6897,
      "step": 862
    },
    {
      "epoch": 0.1015413578068008,
      "grad_norm": 6.424959659576416,
      "learning_rate": 1e-06,
      "loss": 0.7233,
      "step": 863
    },
    {
      "epoch": 0.1016590187080833,
      "grad_norm": 5.144840240478516,
      "learning_rate": 1e-06,
      "loss": 0.6463,
      "step": 864
    },
    {
      "epoch": 0.1017766796093658,
      "grad_norm": 4.471700668334961,
      "learning_rate": 1e-06,
      "loss": 0.7613,
      "step": 865
    },
    {
      "epoch": 0.10189434051064832,
      "grad_norm": 3.5130629539489746,
      "learning_rate": 1e-06,
      "loss": 0.6285,
      "step": 866
    },
    {
      "epoch": 0.10201200141193081,
      "grad_norm": 6.073313236236572,
      "learning_rate": 1e-06,
      "loss": 0.7526,
      "step": 867
    },
    {
      "epoch": 0.10212966231321331,
      "grad_norm": 5.274708271026611,
      "learning_rate": 1e-06,
      "loss": 0.694,
      "step": 868
    },
    {
      "epoch": 0.10224732321449582,
      "grad_norm": 5.783721446990967,
      "learning_rate": 1e-06,
      "loss": 0.7246,
      "step": 869
    },
    {
      "epoch": 0.10236498411577832,
      "grad_norm": 2.880633592605591,
      "learning_rate": 1e-06,
      "loss": 0.7076,
      "step": 870
    },
    {
      "epoch": 0.10248264501706084,
      "grad_norm": 4.704663276672363,
      "learning_rate": 1e-06,
      "loss": 0.7082,
      "step": 871
    },
    {
      "epoch": 0.10260030591834333,
      "grad_norm": 3.2478537559509277,
      "learning_rate": 1e-06,
      "loss": 0.7734,
      "step": 872
    },
    {
      "epoch": 0.10271796681962583,
      "grad_norm": 18.305561065673828,
      "learning_rate": 1e-06,
      "loss": 0.6345,
      "step": 873
    },
    {
      "epoch": 0.10283562772090835,
      "grad_norm": 6.261775016784668,
      "learning_rate": 1e-06,
      "loss": 0.6628,
      "step": 874
    },
    {
      "epoch": 0.10295328862219084,
      "grad_norm": 6.447326183319092,
      "learning_rate": 1e-06,
      "loss": 0.8169,
      "step": 875
    },
    {
      "epoch": 0.10307094952347334,
      "grad_norm": 3.332118511199951,
      "learning_rate": 1e-06,
      "loss": 0.7389,
      "step": 876
    },
    {
      "epoch": 0.10318861042475586,
      "grad_norm": 7.243100166320801,
      "learning_rate": 1e-06,
      "loss": 0.8797,
      "step": 877
    },
    {
      "epoch": 0.10330627132603835,
      "grad_norm": 3.4968297481536865,
      "learning_rate": 1e-06,
      "loss": 0.6324,
      "step": 878
    },
    {
      "epoch": 0.10342393222732087,
      "grad_norm": 3.1249096393585205,
      "learning_rate": 1e-06,
      "loss": 0.7032,
      "step": 879
    },
    {
      "epoch": 0.10354159312860337,
      "grad_norm": 4.255460262298584,
      "learning_rate": 1e-06,
      "loss": 0.8448,
      "step": 880
    },
    {
      "epoch": 0.10365925402988586,
      "grad_norm": 3.519742488861084,
      "learning_rate": 1e-06,
      "loss": 0.7932,
      "step": 881
    },
    {
      "epoch": 0.10377691493116838,
      "grad_norm": 28.501571655273438,
      "learning_rate": 1e-06,
      "loss": 0.696,
      "step": 882
    },
    {
      "epoch": 0.10389457583245088,
      "grad_norm": 2.5168793201446533,
      "learning_rate": 1e-06,
      "loss": 0.881,
      "step": 883
    },
    {
      "epoch": 0.10401223673373337,
      "grad_norm": 8.99081039428711,
      "learning_rate": 1e-06,
      "loss": 0.7411,
      "step": 884
    },
    {
      "epoch": 0.10412989763501589,
      "grad_norm": 13.516379356384277,
      "learning_rate": 1e-06,
      "loss": 0.608,
      "step": 885
    },
    {
      "epoch": 0.10424755853629838,
      "grad_norm": 4.074734687805176,
      "learning_rate": 1e-06,
      "loss": 0.7659,
      "step": 886
    },
    {
      "epoch": 0.1043652194375809,
      "grad_norm": 2.260159492492676,
      "learning_rate": 1e-06,
      "loss": 0.8207,
      "step": 887
    },
    {
      "epoch": 0.1044828803388634,
      "grad_norm": 2.8651790618896484,
      "learning_rate": 1e-06,
      "loss": 0.7703,
      "step": 888
    },
    {
      "epoch": 0.1046005412401459,
      "grad_norm": 4.478638648986816,
      "learning_rate": 1e-06,
      "loss": 0.775,
      "step": 889
    },
    {
      "epoch": 0.10471820214142841,
      "grad_norm": 3.101065158843994,
      "learning_rate": 1e-06,
      "loss": 0.6703,
      "step": 890
    },
    {
      "epoch": 0.1048358630427109,
      "grad_norm": 3.55016827583313,
      "learning_rate": 1e-06,
      "loss": 0.7054,
      "step": 891
    },
    {
      "epoch": 0.1049535239439934,
      "grad_norm": 4.5212507247924805,
      "learning_rate": 1e-06,
      "loss": 0.7758,
      "step": 892
    },
    {
      "epoch": 0.10507118484527592,
      "grad_norm": 4.663824081420898,
      "learning_rate": 1e-06,
      "loss": 0.6694,
      "step": 893
    },
    {
      "epoch": 0.10518884574655842,
      "grad_norm": 4.49483585357666,
      "learning_rate": 1e-06,
      "loss": 0.7006,
      "step": 894
    },
    {
      "epoch": 0.10530650664784093,
      "grad_norm": 2.8132238388061523,
      "learning_rate": 1e-06,
      "loss": 0.8139,
      "step": 895
    },
    {
      "epoch": 0.10542416754912343,
      "grad_norm": 5.475927352905273,
      "learning_rate": 1e-06,
      "loss": 0.7546,
      "step": 896
    },
    {
      "epoch": 0.10554182845040593,
      "grad_norm": 2.9564592838287354,
      "learning_rate": 1e-06,
      "loss": 0.7466,
      "step": 897
    },
    {
      "epoch": 0.10565948935168844,
      "grad_norm": 3.7339460849761963,
      "learning_rate": 1e-06,
      "loss": 0.6938,
      "step": 898
    },
    {
      "epoch": 0.10577715025297094,
      "grad_norm": 4.033295631408691,
      "learning_rate": 1e-06,
      "loss": 0.7168,
      "step": 899
    },
    {
      "epoch": 0.10589481115425343,
      "grad_norm": 3.160860300064087,
      "learning_rate": 1e-06,
      "loss": 0.6834,
      "step": 900
    },
    {
      "epoch": 0.10601247205553595,
      "grad_norm": 7.493616104125977,
      "learning_rate": 1e-06,
      "loss": 0.7393,
      "step": 901
    },
    {
      "epoch": 0.10613013295681845,
      "grad_norm": 22.620590209960938,
      "learning_rate": 1e-06,
      "loss": 0.8435,
      "step": 902
    },
    {
      "epoch": 0.10624779385810096,
      "grad_norm": 3.6928799152374268,
      "learning_rate": 1e-06,
      "loss": 0.7645,
      "step": 903
    },
    {
      "epoch": 0.10636545475938346,
      "grad_norm": 3.0258662700653076,
      "learning_rate": 1e-06,
      "loss": 0.6949,
      "step": 904
    },
    {
      "epoch": 0.10648311566066596,
      "grad_norm": 7.960806846618652,
      "learning_rate": 1e-06,
      "loss": 0.6814,
      "step": 905
    },
    {
      "epoch": 0.10660077656194847,
      "grad_norm": 10.985957145690918,
      "learning_rate": 1e-06,
      "loss": 0.8068,
      "step": 906
    },
    {
      "epoch": 0.10671843746323097,
      "grad_norm": 5.588890075683594,
      "learning_rate": 1e-06,
      "loss": 0.712,
      "step": 907
    },
    {
      "epoch": 0.10683609836451347,
      "grad_norm": 2.6326663494110107,
      "learning_rate": 1e-06,
      "loss": 0.8733,
      "step": 908
    },
    {
      "epoch": 0.10695375926579598,
      "grad_norm": 6.32219123840332,
      "learning_rate": 1e-06,
      "loss": 0.7209,
      "step": 909
    },
    {
      "epoch": 0.10707142016707848,
      "grad_norm": 3.7147603034973145,
      "learning_rate": 1e-06,
      "loss": 0.6518,
      "step": 910
    },
    {
      "epoch": 0.10718908106836099,
      "grad_norm": 6.2384352684021,
      "learning_rate": 1e-06,
      "loss": 0.7457,
      "step": 911
    },
    {
      "epoch": 0.10730674196964349,
      "grad_norm": 6.7992987632751465,
      "learning_rate": 1e-06,
      "loss": 0.6877,
      "step": 912
    },
    {
      "epoch": 0.10742440287092599,
      "grad_norm": 4.401557922363281,
      "learning_rate": 1e-06,
      "loss": 0.7824,
      "step": 913
    },
    {
      "epoch": 0.1075420637722085,
      "grad_norm": 4.7830705642700195,
      "learning_rate": 1e-06,
      "loss": 0.7996,
      "step": 914
    },
    {
      "epoch": 0.107659724673491,
      "grad_norm": 5.023268699645996,
      "learning_rate": 1e-06,
      "loss": 0.8633,
      "step": 915
    },
    {
      "epoch": 0.1077773855747735,
      "grad_norm": 5.523383617401123,
      "learning_rate": 1e-06,
      "loss": 0.8391,
      "step": 916
    },
    {
      "epoch": 0.10789504647605601,
      "grad_norm": 3.5851926803588867,
      "learning_rate": 1e-06,
      "loss": 0.8981,
      "step": 917
    },
    {
      "epoch": 0.10801270737733851,
      "grad_norm": 6.372958183288574,
      "learning_rate": 1e-06,
      "loss": 0.7296,
      "step": 918
    },
    {
      "epoch": 0.10813036827862102,
      "grad_norm": 3.6867308616638184,
      "learning_rate": 1e-06,
      "loss": 0.8342,
      "step": 919
    },
    {
      "epoch": 0.10824802917990352,
      "grad_norm": 3.857619047164917,
      "learning_rate": 1e-06,
      "loss": 0.6729,
      "step": 920
    },
    {
      "epoch": 0.10836569008118602,
      "grad_norm": 10.918023109436035,
      "learning_rate": 1e-06,
      "loss": 0.7497,
      "step": 921
    },
    {
      "epoch": 0.10848335098246853,
      "grad_norm": 4.183814525604248,
      "learning_rate": 1e-06,
      "loss": 0.7404,
      "step": 922
    },
    {
      "epoch": 0.10860101188375103,
      "grad_norm": 6.011120796203613,
      "learning_rate": 1e-06,
      "loss": 0.7812,
      "step": 923
    },
    {
      "epoch": 0.10871867278503353,
      "grad_norm": 3.549236297607422,
      "learning_rate": 1e-06,
      "loss": 0.7345,
      "step": 924
    },
    {
      "epoch": 0.10883633368631604,
      "grad_norm": 3.6891543865203857,
      "learning_rate": 1e-06,
      "loss": 0.7796,
      "step": 925
    },
    {
      "epoch": 0.10895399458759854,
      "grad_norm": 7.792111396789551,
      "learning_rate": 1e-06,
      "loss": 0.7652,
      "step": 926
    },
    {
      "epoch": 0.10907165548888105,
      "grad_norm": 4.561732769012451,
      "learning_rate": 1e-06,
      "loss": 0.7025,
      "step": 927
    },
    {
      "epoch": 0.10918931639016355,
      "grad_norm": 5.012827396392822,
      "learning_rate": 1e-06,
      "loss": 0.7675,
      "step": 928
    },
    {
      "epoch": 0.10930697729144605,
      "grad_norm": 4.815629959106445,
      "learning_rate": 1e-06,
      "loss": 0.7956,
      "step": 929
    },
    {
      "epoch": 0.10942463819272856,
      "grad_norm": 7.950117588043213,
      "learning_rate": 1e-06,
      "loss": 0.7514,
      "step": 930
    },
    {
      "epoch": 0.10954229909401106,
      "grad_norm": 3.6295828819274902,
      "learning_rate": 1e-06,
      "loss": 0.7498,
      "step": 931
    },
    {
      "epoch": 0.10965995999529356,
      "grad_norm": 8.010412216186523,
      "learning_rate": 1e-06,
      "loss": 0.6758,
      "step": 932
    },
    {
      "epoch": 0.10977762089657607,
      "grad_norm": 3.3440520763397217,
      "learning_rate": 1e-06,
      "loss": 0.8549,
      "step": 933
    },
    {
      "epoch": 0.10989528179785857,
      "grad_norm": 3.4136955738067627,
      "learning_rate": 1e-06,
      "loss": 0.693,
      "step": 934
    },
    {
      "epoch": 0.11001294269914108,
      "grad_norm": 4.733597755432129,
      "learning_rate": 1e-06,
      "loss": 0.8341,
      "step": 935
    },
    {
      "epoch": 0.11013060360042358,
      "grad_norm": 13.338993072509766,
      "learning_rate": 1e-06,
      "loss": 0.7849,
      "step": 936
    },
    {
      "epoch": 0.11024826450170608,
      "grad_norm": 3.829407215118408,
      "learning_rate": 1e-06,
      "loss": 0.7122,
      "step": 937
    },
    {
      "epoch": 0.11036592540298859,
      "grad_norm": 5.739267826080322,
      "learning_rate": 1e-06,
      "loss": 0.7708,
      "step": 938
    },
    {
      "epoch": 0.11048358630427109,
      "grad_norm": 5.647058010101318,
      "learning_rate": 1e-06,
      "loss": 0.7333,
      "step": 939
    },
    {
      "epoch": 0.11060124720555359,
      "grad_norm": 3.6504688262939453,
      "learning_rate": 1e-06,
      "loss": 0.678,
      "step": 940
    },
    {
      "epoch": 0.1107189081068361,
      "grad_norm": 4.032068729400635,
      "learning_rate": 1e-06,
      "loss": 0.6942,
      "step": 941
    },
    {
      "epoch": 0.1108365690081186,
      "grad_norm": 3.7702181339263916,
      "learning_rate": 1e-06,
      "loss": 0.7686,
      "step": 942
    },
    {
      "epoch": 0.11095422990940111,
      "grad_norm": 10.096896171569824,
      "learning_rate": 1e-06,
      "loss": 0.8725,
      "step": 943
    },
    {
      "epoch": 0.11107189081068361,
      "grad_norm": 5.022398948669434,
      "learning_rate": 1e-06,
      "loss": 0.7321,
      "step": 944
    },
    {
      "epoch": 0.11118955171196611,
      "grad_norm": 3.7425436973571777,
      "learning_rate": 1e-06,
      "loss": 0.6971,
      "step": 945
    },
    {
      "epoch": 0.11130721261324862,
      "grad_norm": 17.83504295349121,
      "learning_rate": 1e-06,
      "loss": 0.7606,
      "step": 946
    },
    {
      "epoch": 0.11142487351453112,
      "grad_norm": 5.255575180053711,
      "learning_rate": 1e-06,
      "loss": 0.7614,
      "step": 947
    },
    {
      "epoch": 0.11154253441581362,
      "grad_norm": 6.864058494567871,
      "learning_rate": 1e-06,
      "loss": 0.7047,
      "step": 948
    },
    {
      "epoch": 0.11166019531709613,
      "grad_norm": 8.533795356750488,
      "learning_rate": 1e-06,
      "loss": 0.8295,
      "step": 949
    },
    {
      "epoch": 0.11177785621837863,
      "grad_norm": 8.808038711547852,
      "learning_rate": 1e-06,
      "loss": 0.81,
      "step": 950
    },
    {
      "epoch": 0.11189551711966114,
      "grad_norm": 3.7528939247131348,
      "learning_rate": 1e-06,
      "loss": 0.6968,
      "step": 951
    },
    {
      "epoch": 0.11201317802094364,
      "grad_norm": 12.15597915649414,
      "learning_rate": 1e-06,
      "loss": 0.6969,
      "step": 952
    },
    {
      "epoch": 0.11213083892222614,
      "grad_norm": 3.9877498149871826,
      "learning_rate": 1e-06,
      "loss": 0.7565,
      "step": 953
    },
    {
      "epoch": 0.11224849982350865,
      "grad_norm": 3.230027914047241,
      "learning_rate": 1e-06,
      "loss": 0.7655,
      "step": 954
    },
    {
      "epoch": 0.11236616072479115,
      "grad_norm": 5.134922504425049,
      "learning_rate": 1e-06,
      "loss": 0.8121,
      "step": 955
    },
    {
      "epoch": 0.11248382162607365,
      "grad_norm": 3.7903997898101807,
      "learning_rate": 1e-06,
      "loss": 0.7504,
      "step": 956
    },
    {
      "epoch": 0.11260148252735616,
      "grad_norm": 3.913166046142578,
      "learning_rate": 1e-06,
      "loss": 0.7001,
      "step": 957
    },
    {
      "epoch": 0.11271914342863866,
      "grad_norm": 5.1488165855407715,
      "learning_rate": 1e-06,
      "loss": 0.7735,
      "step": 958
    },
    {
      "epoch": 0.11283680432992117,
      "grad_norm": 3.0253422260284424,
      "learning_rate": 1e-06,
      "loss": 0.8549,
      "step": 959
    },
    {
      "epoch": 0.11295446523120367,
      "grad_norm": 15.515777587890625,
      "learning_rate": 1e-06,
      "loss": 0.7316,
      "step": 960
    },
    {
      "epoch": 0.11307212613248617,
      "grad_norm": 4.447922706604004,
      "learning_rate": 1e-06,
      "loss": 0.802,
      "step": 961
    },
    {
      "epoch": 0.11318978703376868,
      "grad_norm": 6.39625883102417,
      "learning_rate": 1e-06,
      "loss": 0.7921,
      "step": 962
    },
    {
      "epoch": 0.11330744793505118,
      "grad_norm": 10.484755516052246,
      "learning_rate": 1e-06,
      "loss": 0.7821,
      "step": 963
    },
    {
      "epoch": 0.11342510883633368,
      "grad_norm": 10.59896469116211,
      "learning_rate": 1e-06,
      "loss": 0.7996,
      "step": 964
    },
    {
      "epoch": 0.11354276973761619,
      "grad_norm": 6.321623802185059,
      "learning_rate": 1e-06,
      "loss": 0.8076,
      "step": 965
    },
    {
      "epoch": 0.11366043063889869,
      "grad_norm": 2.905377149581909,
      "learning_rate": 1e-06,
      "loss": 0.5892,
      "step": 966
    },
    {
      "epoch": 0.1137780915401812,
      "grad_norm": 4.772261619567871,
      "learning_rate": 1e-06,
      "loss": 0.7141,
      "step": 967
    },
    {
      "epoch": 0.1138957524414637,
      "grad_norm": 4.624825954437256,
      "learning_rate": 1e-06,
      "loss": 0.6739,
      "step": 968
    },
    {
      "epoch": 0.1140134133427462,
      "grad_norm": 9.244427680969238,
      "learning_rate": 1e-06,
      "loss": 0.7439,
      "step": 969
    },
    {
      "epoch": 0.11413107424402871,
      "grad_norm": 4.125616550445557,
      "learning_rate": 1e-06,
      "loss": 0.8356,
      "step": 970
    },
    {
      "epoch": 0.11424873514531121,
      "grad_norm": 2.889173746109009,
      "learning_rate": 1e-06,
      "loss": 0.751,
      "step": 971
    },
    {
      "epoch": 0.11436639604659371,
      "grad_norm": 4.325053691864014,
      "learning_rate": 1e-06,
      "loss": 0.8099,
      "step": 972
    },
    {
      "epoch": 0.11448405694787622,
      "grad_norm": 3.6954429149627686,
      "learning_rate": 1e-06,
      "loss": 0.9103,
      "step": 973
    },
    {
      "epoch": 0.11460171784915872,
      "grad_norm": 3.570265293121338,
      "learning_rate": 1e-06,
      "loss": 0.7352,
      "step": 974
    },
    {
      "epoch": 0.11471937875044123,
      "grad_norm": 3.713390350341797,
      "learning_rate": 1e-06,
      "loss": 0.6956,
      "step": 975
    },
    {
      "epoch": 0.11483703965172373,
      "grad_norm": 6.262350559234619,
      "learning_rate": 1e-06,
      "loss": 0.7834,
      "step": 976
    },
    {
      "epoch": 0.11495470055300623,
      "grad_norm": 13.737163543701172,
      "learning_rate": 1e-06,
      "loss": 0.6825,
      "step": 977
    },
    {
      "epoch": 0.11507236145428874,
      "grad_norm": 3.9696526527404785,
      "learning_rate": 1e-06,
      "loss": 0.8195,
      "step": 978
    },
    {
      "epoch": 0.11519002235557124,
      "grad_norm": 3.6462998390197754,
      "learning_rate": 1e-06,
      "loss": 0.8344,
      "step": 979
    },
    {
      "epoch": 0.11530768325685374,
      "grad_norm": 3.0372889041900635,
      "learning_rate": 1e-06,
      "loss": 0.8347,
      "step": 980
    },
    {
      "epoch": 0.11542534415813625,
      "grad_norm": 2.526247501373291,
      "learning_rate": 1e-06,
      "loss": 0.9124,
      "step": 981
    },
    {
      "epoch": 0.11554300505941875,
      "grad_norm": 3.632237195968628,
      "learning_rate": 1e-06,
      "loss": 0.6517,
      "step": 982
    },
    {
      "epoch": 0.11566066596070126,
      "grad_norm": 3.3434395790100098,
      "learning_rate": 1e-06,
      "loss": 0.7136,
      "step": 983
    },
    {
      "epoch": 0.11577832686198376,
      "grad_norm": 4.603308200836182,
      "learning_rate": 1e-06,
      "loss": 0.8149,
      "step": 984
    },
    {
      "epoch": 0.11589598776326626,
      "grad_norm": 4.951627731323242,
      "learning_rate": 1e-06,
      "loss": 0.7847,
      "step": 985
    },
    {
      "epoch": 0.11601364866454877,
      "grad_norm": 4.969621658325195,
      "learning_rate": 1e-06,
      "loss": 0.7254,
      "step": 986
    },
    {
      "epoch": 0.11613130956583127,
      "grad_norm": 3.2793381214141846,
      "learning_rate": 1e-06,
      "loss": 0.7147,
      "step": 987
    },
    {
      "epoch": 0.11624897046711377,
      "grad_norm": 2.3934788703918457,
      "learning_rate": 1e-06,
      "loss": 0.7056,
      "step": 988
    },
    {
      "epoch": 0.11636663136839628,
      "grad_norm": 3.8675055503845215,
      "learning_rate": 1e-06,
      "loss": 0.8171,
      "step": 989
    },
    {
      "epoch": 0.11648429226967878,
      "grad_norm": 5.718388080596924,
      "learning_rate": 1e-06,
      "loss": 0.813,
      "step": 990
    },
    {
      "epoch": 0.1166019531709613,
      "grad_norm": 4.406123161315918,
      "learning_rate": 1e-06,
      "loss": 0.8078,
      "step": 991
    },
    {
      "epoch": 0.1167196140722438,
      "grad_norm": 2.9738152027130127,
      "learning_rate": 1e-06,
      "loss": 0.7428,
      "step": 992
    },
    {
      "epoch": 0.11683727497352629,
      "grad_norm": 6.35261344909668,
      "learning_rate": 1e-06,
      "loss": 0.7331,
      "step": 993
    },
    {
      "epoch": 0.1169549358748088,
      "grad_norm": 6.00268030166626,
      "learning_rate": 1e-06,
      "loss": 0.766,
      "step": 994
    },
    {
      "epoch": 0.1170725967760913,
      "grad_norm": 12.03709888458252,
      "learning_rate": 1e-06,
      "loss": 0.7315,
      "step": 995
    },
    {
      "epoch": 0.1171902576773738,
      "grad_norm": 6.83236837387085,
      "learning_rate": 1e-06,
      "loss": 0.7091,
      "step": 996
    },
    {
      "epoch": 0.11730791857865631,
      "grad_norm": 6.966513633728027,
      "learning_rate": 1e-06,
      "loss": 0.8094,
      "step": 997
    },
    {
      "epoch": 0.11742557947993881,
      "grad_norm": 5.393833637237549,
      "learning_rate": 1e-06,
      "loss": 0.6795,
      "step": 998
    },
    {
      "epoch": 0.11754324038122133,
      "grad_norm": 4.947326183319092,
      "learning_rate": 1e-06,
      "loss": 0.8499,
      "step": 999
    },
    {
      "epoch": 0.11766090128250382,
      "grad_norm": 3.890005350112915,
      "learning_rate": 1e-06,
      "loss": 0.7317,
      "step": 1000
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.6569126454516777e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}