{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 1004,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00796812749003984,
      "grad_norm": 1.5820205211639404,
      "learning_rate": 9.803921568627452e-08,
      "loss": 2.0908608436584473,
      "step": 2
    },
    {
      "epoch": 0.01593625498007968,
      "grad_norm": 0.6021012663841248,
      "learning_rate": 2.9411764705882356e-07,
      "loss": 2.011705160140991,
      "step": 4
    },
    {
      "epoch": 0.02390438247011952,
      "grad_norm": 2.651017904281616,
      "learning_rate": 4.901960784313725e-07,
      "loss": 2.2025299072265625,
      "step": 6
    },
    {
      "epoch": 0.03187250996015936,
      "grad_norm": 0.6893859505653381,
      "learning_rate": 6.862745098039217e-07,
      "loss": 1.915262222290039,
      "step": 8
    },
    {
      "epoch": 0.0398406374501992,
      "grad_norm": 5.2249884605407715,
      "learning_rate": 8.823529411764707e-07,
      "loss": 1.5706487894058228,
      "step": 10
    },
    {
      "epoch": 0.04780876494023904,
      "grad_norm": 0.9423531889915466,
      "learning_rate": 1.0784313725490197e-06,
      "loss": 1.8168272972106934,
      "step": 12
    },
    {
      "epoch": 0.055776892430278883,
      "grad_norm": 0.6569095849990845,
      "learning_rate": 1.2745098039215686e-06,
      "loss": 2.7349138259887695,
      "step": 14
    },
    {
      "epoch": 0.06374501992031872,
      "grad_norm": 3.7171003818511963,
      "learning_rate": 1.4705882352941177e-06,
      "loss": 4.141521453857422,
      "step": 16
    },
    {
      "epoch": 0.07171314741035857,
      "grad_norm": 0.8796786665916443,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.7450640201568604,
      "step": 18
    },
    {
      "epoch": 0.0796812749003984,
      "grad_norm": 0.5898110270500183,
      "learning_rate": 1.8627450980392158e-06,
      "loss": 1.646957278251648,
      "step": 20
    },
    {
      "epoch": 0.08764940239043825,
      "grad_norm": 0.6275415420532227,
      "learning_rate": 2.058823529411765e-06,
      "loss": 1.5267493724822998,
      "step": 22
    },
    {
      "epoch": 0.09561752988047809,
      "grad_norm": 0.994874894618988,
      "learning_rate": 2.254901960784314e-06,
      "loss": 1.4252383708953857,
      "step": 24
    },
    {
      "epoch": 0.10358565737051793,
      "grad_norm": 0.33096620440483093,
      "learning_rate": 2.450980392156863e-06,
      "loss": 1.5317208766937256,
      "step": 26
    },
    {
      "epoch": 0.11155378486055777,
      "grad_norm": 0.5782793760299683,
      "learning_rate": 2.647058823529412e-06,
      "loss": 1.6394634246826172,
      "step": 28
    },
    {
      "epoch": 0.11952191235059761,
      "grad_norm": 1.350926160812378,
      "learning_rate": 2.843137254901961e-06,
      "loss": 1.3870218992233276,
      "step": 30
    },
    {
      "epoch": 0.12749003984063745,
      "grad_norm": 0.7801507711410522,
      "learning_rate": 3.03921568627451e-06,
      "loss": 1.206599235534668,
      "step": 32
    },
    {
      "epoch": 0.13545816733067728,
      "grad_norm": 1.6924139261245728,
      "learning_rate": 3.2352941176470594e-06,
      "loss": 1.2964345216751099,
      "step": 34
    },
    {
      "epoch": 0.14342629482071714,
      "grad_norm": 0.38787487149238586,
      "learning_rate": 3.431372549019608e-06,
      "loss": 1.4511497020721436,
      "step": 36
    },
    {
      "epoch": 0.15139442231075698,
      "grad_norm": 0.6149557828903198,
      "learning_rate": 3.6274509803921573e-06,
      "loss": 1.228920817375183,
      "step": 38
    },
    {
      "epoch": 0.1593625498007968,
      "grad_norm": 0.3690328299999237,
      "learning_rate": 3.8235294117647055e-06,
      "loss": 1.357020378112793,
      "step": 40
    },
    {
      "epoch": 0.16733067729083664,
      "grad_norm": 0.8299920558929443,
      "learning_rate": 4.019607843137255e-06,
      "loss": 0.8516656756401062,
      "step": 42
    },
    {
      "epoch": 0.1752988047808765,
      "grad_norm": 1.2672408819198608,
      "learning_rate": 4.215686274509805e-06,
      "loss": 1.0402929782867432,
      "step": 44
    },
    {
      "epoch": 0.18326693227091634,
      "grad_norm": 0.4577001631259918,
      "learning_rate": 4.411764705882353e-06,
      "loss": 0.9601466655731201,
      "step": 46
    },
    {
      "epoch": 0.19123505976095617,
      "grad_norm": 0.5572270750999451,
      "learning_rate": 4.607843137254902e-06,
      "loss": 1.206835389137268,
      "step": 48
    },
    {
      "epoch": 0.199203187250996,
      "grad_norm": 0.4704183042049408,
      "learning_rate": 4.803921568627452e-06,
      "loss": 1.3656173944473267,
      "step": 50
    },
    {
      "epoch": 0.20717131474103587,
      "grad_norm": 0.3990301191806793,
      "learning_rate": 5e-06,
      "loss": 1.3112876415252686,
      "step": 52
    },
    {
      "epoch": 0.2151394422310757,
      "grad_norm": 3.5827512741088867,
      "learning_rate": 4.999951098195543e-06,
      "loss": 1.1392691135406494,
      "step": 54
    },
    {
      "epoch": 0.22310756972111553,
      "grad_norm": 0.9888070821762085,
      "learning_rate": 4.9998043949078495e-06,
      "loss": 1.5367088317871094,
      "step": 56
    },
    {
      "epoch": 0.23107569721115537,
      "grad_norm": 0.5973958373069763,
      "learning_rate": 4.9995598965138565e-06,
      "loss": 1.3594309091567993,
      "step": 58
    },
    {
      "epoch": 0.23904382470119523,
      "grad_norm": 0.4080948531627655,
      "learning_rate": 4.9992176136414875e-06,
      "loss": 0.7706686854362488,
      "step": 60
    },
    {
      "epoch": 0.24701195219123506,
      "grad_norm": 0.3561408221721649,
      "learning_rate": 4.998777561169186e-06,
      "loss": 1.3637105226516724,
      "step": 62
    },
    {
      "epoch": 0.2549800796812749,
      "grad_norm": 0.7727970480918884,
      "learning_rate": 4.998239758225272e-06,
      "loss": 1.123347282409668,
      "step": 64
    },
    {
      "epoch": 0.26294820717131473,
      "grad_norm": 0.454253226518631,
      "learning_rate": 4.99760422818711e-06,
      "loss": 1.2842735052108765,
      "step": 66
    },
    {
      "epoch": 0.27091633466135456,
      "grad_norm": 0.8649628758430481,
      "learning_rate": 4.9968709986800925e-06,
      "loss": 0.7890381217002869,
      "step": 68
    },
    {
      "epoch": 0.2788844621513944,
      "grad_norm": 0.5761728882789612,
      "learning_rate": 4.996040101576436e-06,
      "loss": 1.5591208934783936,
      "step": 70
    },
    {
      "epoch": 0.2868525896414343,
      "grad_norm": 0.7316920161247253,
      "learning_rate": 4.995111572993804e-06,
      "loss": 0.808398962020874,
      "step": 72
    },
    {
      "epoch": 0.2948207171314741,
      "grad_norm": 0.6703289151191711,
      "learning_rate": 4.994085453293725e-06,
      "loss": 1.1256080865859985,
      "step": 74
    },
    {
      "epoch": 0.30278884462151395,
      "grad_norm": 0.747173011302948,
      "learning_rate": 4.992961787079848e-06,
      "loss": 1.4557641744613647,
      "step": 76
    },
    {
      "epoch": 0.3107569721115538,
      "grad_norm": 1.3619074821472168,
      "learning_rate": 4.991740623196001e-06,
      "loss": 0.8900258541107178,
      "step": 78
    },
    {
      "epoch": 0.3187250996015936,
      "grad_norm": 0.2858664095401764,
      "learning_rate": 4.990422014724068e-06,
      "loss": 0.8188052177429199,
      "step": 80
    },
    {
      "epoch": 0.32669322709163345,
      "grad_norm": 0.7998085618019104,
      "learning_rate": 4.989006018981675e-06,
      "loss": 1.3202816247940063,
      "step": 82
    },
    {
      "epoch": 0.3346613545816733,
      "grad_norm": 0.7839249968528748,
      "learning_rate": 4.987492697519712e-06,
      "loss": 0.8596320152282715,
      "step": 84
    },
    {
      "epoch": 0.3426294820717131,
      "grad_norm": 0.602087676525116,
      "learning_rate": 4.985882116119643e-06,
      "loss": 1.2963725328445435,
      "step": 86
    },
    {
      "epoch": 0.350597609561753,
      "grad_norm": 0.730972945690155,
      "learning_rate": 4.9841743447906576e-06,
      "loss": 0.90030837059021,
      "step": 88
    },
    {
      "epoch": 0.35856573705179284,
      "grad_norm": 0.4549611508846283,
      "learning_rate": 4.982369457766621e-06,
      "loss": 1.3182578086853027,
      "step": 90
    },
    {
      "epoch": 0.3665338645418327,
      "grad_norm": 0.7147920727729797,
      "learning_rate": 4.980467533502853e-06,
      "loss": 1.1057542562484741,
      "step": 92
    },
    {
      "epoch": 0.3745019920318725,
      "grad_norm": 0.5205848217010498,
      "learning_rate": 4.978468654672709e-06,
      "loss": 1.2402427196502686,
      "step": 94
    },
    {
      "epoch": 0.38247011952191234,
      "grad_norm": 5.3043951988220215,
      "learning_rate": 4.976372908163999e-06,
      "loss": 1.0836384296417236,
      "step": 96
    },
    {
      "epoch": 0.3904382470119522,
      "grad_norm": 0.5307750105857849,
      "learning_rate": 4.974180385075195e-06,
      "loss": 1.2550530433654785,
      "step": 98
    },
    {
      "epoch": 0.398406374501992,
      "grad_norm": 0.46418726444244385,
      "learning_rate": 4.971891180711488e-06,
      "loss": 0.8554410934448242,
      "step": 100
    },
    {
      "epoch": 0.4063745019920319,
      "grad_norm": 0.48276621103286743,
      "learning_rate": 4.9695053945806295e-06,
      "loss": 1.266715407371521,
      "step": 102
    },
    {
      "epoch": 0.41434262948207173,
      "grad_norm": 0.6375750303268433,
      "learning_rate": 4.9670231303886194e-06,
      "loss": 0.8264124393463135,
      "step": 104
    },
    {
      "epoch": 0.42231075697211157,
      "grad_norm": 1.626670241355896,
      "learning_rate": 4.9644444960351865e-06,
      "loss": 0.7509417533874512,
      "step": 106
    },
    {
      "epoch": 0.4302788844621514,
      "grad_norm": 0.504277229309082,
      "learning_rate": 4.9617696036091095e-06,
      "loss": 1.0068732500076294,
      "step": 108
    },
    {
      "epoch": 0.43824701195219123,
      "grad_norm": 0.464828759431839,
      "learning_rate": 4.958998569383335e-06,
      "loss": 1.2774951457977295,
      "step": 110
    },
    {
      "epoch": 0.44621513944223107,
      "grad_norm": 0.7254738807678223,
      "learning_rate": 4.956131513809928e-06,
      "loss": 0.8106322884559631,
      "step": 112
    },
    {
      "epoch": 0.4541832669322709,
      "grad_norm": 0.6553805470466614,
      "learning_rate": 4.953168561514838e-06,
      "loss": 1.1187655925750732,
      "step": 114
    },
    {
      "epoch": 0.46215139442231074,
      "grad_norm": 0.48421838879585266,
      "learning_rate": 4.950109841292475e-06,
      "loss": 1.2787994146347046,
      "step": 116
    },
    {
      "epoch": 0.4701195219123506,
      "grad_norm": 0.38402482867240906,
      "learning_rate": 4.946955486100119e-06,
      "loss": 1.3395017385482788,
      "step": 118
    },
    {
      "epoch": 0.47808764940239046,
      "grad_norm": 0.3643367886543274,
      "learning_rate": 4.943705633052134e-06,
      "loss": 1.2211538553237915,
      "step": 120
    },
    {
      "epoch": 0.4860557768924303,
      "grad_norm": 0.6005005836486816,
      "learning_rate": 4.940360423414013e-06,
      "loss": 0.9491733908653259,
      "step": 122
    },
    {
      "epoch": 0.4940239043824701,
      "grad_norm": 1.9367928504943848,
      "learning_rate": 4.936920002596233e-06,
      "loss": 1.0181323289871216,
      "step": 124
    },
    {
      "epoch": 0.50199203187251,
      "grad_norm": 0.9496893286705017,
      "learning_rate": 4.9333845201479375e-06,
      "loss": 0.9098525643348694,
      "step": 126
    },
    {
      "epoch": 0.5099601593625498,
      "grad_norm": 0.3858175575733185,
      "learning_rate": 4.9297541297504345e-06,
      "loss": 1.2625067234039307,
      "step": 128
    },
    {
      "epoch": 0.5179282868525896,
      "grad_norm": 0.6485924124717712,
      "learning_rate": 4.9260289892105164e-06,
      "loss": 0.9286548495292664,
      "step": 130
    },
    {
      "epoch": 0.5258964143426295,
      "grad_norm": 0.4828100800514221,
      "learning_rate": 4.922209260453599e-06,
      "loss": 0.9075167775154114,
      "step": 132
    },
    {
      "epoch": 0.5338645418326693,
      "grad_norm": 0.6577417850494385,
      "learning_rate": 4.9182951095166875e-06,
      "loss": 0.9003283381462097,
      "step": 134
    },
    {
      "epoch": 0.5418326693227091,
      "grad_norm": 0.39047494530677795,
      "learning_rate": 4.914286706541153e-06,
      "loss": 1.263765573501587,
      "step": 136
    },
    {
      "epoch": 0.549800796812749,
      "grad_norm": 0.37188729643821716,
      "learning_rate": 4.910184225765341e-06,
      "loss": 0.9224218130111694,
      "step": 138
    },
    {
      "epoch": 0.5577689243027888,
      "grad_norm": 0.700023353099823,
      "learning_rate": 4.905987845516996e-06,
      "loss": 1.0760854482650757,
      "step": 140
    },
    {
      "epoch": 0.5657370517928287,
      "grad_norm": 0.5397536158561707,
      "learning_rate": 4.90169774820551e-06,
      "loss": 1.0052696466445923,
      "step": 142
    },
    {
      "epoch": 0.5737051792828686,
      "grad_norm": 0.3965936005115509,
      "learning_rate": 4.897314120313996e-06,
      "loss": 1.2108368873596191,
      "step": 144
    },
    {
      "epoch": 0.5816733067729084,
      "grad_norm": 0.5652391314506531,
      "learning_rate": 4.892837152391174e-06,
      "loss": 1.3467415571212769,
      "step": 146
    },
    {
      "epoch": 0.5896414342629482,
      "grad_norm": 0.6620450615882874,
      "learning_rate": 4.888267039043101e-06,
      "loss": 0.9482460021972656,
      "step": 148
    },
    {
      "epoch": 0.5976095617529881,
      "grad_norm": 0.9544744491577148,
      "learning_rate": 4.883603978924698e-06,
      "loss": 0.9302203059196472,
      "step": 150
    },
    {
      "epoch": 0.6055776892430279,
      "grad_norm": 2.305940866470337,
      "learning_rate": 4.878848174731126e-06,
      "loss": 0.9413925409317017,
      "step": 152
    },
    {
      "epoch": 0.6135458167330677,
      "grad_norm": 2.283101797103882,
      "learning_rate": 4.873999833188967e-06,
      "loss": 1.1068358421325684,
      "step": 154
    },
    {
      "epoch": 0.6215139442231076,
      "grad_norm": 3.7683253288269043,
      "learning_rate": 4.8690591650472455e-06,
      "loss": 0.9520494341850281,
      "step": 156
    },
    {
      "epoch": 0.6294820717131474,
      "grad_norm": 0.5936416983604431,
      "learning_rate": 4.864026385068259e-06,
      "loss": 1.1352072954177856,
      "step": 158
    },
    {
      "epoch": 0.6374501992031872,
      "grad_norm": 0.4278126060962677,
      "learning_rate": 4.858901712018251e-06,
      "loss": 1.0051214694976807,
      "step": 160
    },
    {
      "epoch": 0.6454183266932271,
      "grad_norm": 0.2299516201019287,
      "learning_rate": 4.8536853686578965e-06,
      "loss": 0.7448744773864746,
      "step": 162
    },
    {
      "epoch": 0.6533864541832669,
      "grad_norm": 0.3577902615070343,
      "learning_rate": 4.84837758173262e-06,
      "loss": 1.1305558681488037,
      "step": 164
    },
    {
      "epoch": 0.6613545816733067,
      "grad_norm": 0.5466222167015076,
      "learning_rate": 4.8429785819627395e-06,
      "loss": 0.7910860776901245,
      "step": 166
    },
    {
      "epoch": 0.6693227091633466,
      "grad_norm": 0.2932773232460022,
      "learning_rate": 4.83748860403344e-06,
      "loss": 1.2195664644241333,
      "step": 168
    },
    {
      "epoch": 0.6772908366533864,
      "grad_norm": 2.732438802719116,
      "learning_rate": 4.831907886584566e-06,
      "loss": 0.8674245476722717,
      "step": 170
    },
    {
      "epoch": 0.6852589641434262,
      "grad_norm": 2.0883586406707764,
      "learning_rate": 4.8262366722002535e-06,
      "loss": 0.813420832157135,
      "step": 172
    },
    {
      "epoch": 0.6932270916334662,
      "grad_norm": 0.3445831835269928,
      "learning_rate": 4.820475207398386e-06,
      "loss": 1.218064546585083,
      "step": 174
    },
    {
      "epoch": 0.701195219123506,
      "grad_norm": 1.5425689220428467,
      "learning_rate": 4.8146237426198735e-06,
      "loss": 0.7231574058532715,
      "step": 176
    },
    {
      "epoch": 0.7091633466135459,
      "grad_norm": 1.0461406707763672,
      "learning_rate": 4.808682532217772e-06,
      "loss": 0.8952961564064026,
      "step": 178
    },
    {
      "epoch": 0.7171314741035857,
      "grad_norm": 1.4359701871871948,
      "learning_rate": 4.802651834446221e-06,
      "loss": 1.0529459714889526,
      "step": 180
    },
    {
      "epoch": 0.7250996015936255,
      "grad_norm": 0.3937617838382721,
      "learning_rate": 4.796531911449226e-06,
      "loss": 0.49386098980903625,
      "step": 182
    },
    {
      "epoch": 0.7330677290836654,
      "grad_norm": 0.28616243600845337,
      "learning_rate": 4.790323029249252e-06,
      "loss": 1.2472022771835327,
      "step": 184
    },
    {
      "epoch": 0.7410358565737052,
      "grad_norm": 0.3717990219593048,
      "learning_rate": 4.784025457735675e-06,
      "loss": 1.229090929031372,
      "step": 186
    },
    {
      "epoch": 0.749003984063745,
      "grad_norm": 0.5032241940498352,
      "learning_rate": 4.777639470653035e-06,
      "loss": 1.0716753005981445,
      "step": 188
    },
    {
      "epoch": 0.7569721115537849,
      "grad_norm": 0.5822710990905762,
      "learning_rate": 4.771165345589148e-06,
      "loss": 0.8541173934936523,
      "step": 190
    },
    {
      "epoch": 0.7649402390438247,
      "grad_norm": 0.621751606464386,
      "learning_rate": 4.7646033639630355e-06,
      "loss": 1.1807188987731934,
      "step": 192
    },
    {
      "epoch": 0.7729083665338645,
      "grad_norm": 1.1548640727996826,
      "learning_rate": 4.757953811012688e-06,
      "loss": 0.8662915229797363,
      "step": 194
    },
    {
      "epoch": 0.7808764940239044,
      "grad_norm": 3.4132649898529053,
      "learning_rate": 4.7512169757826735e-06,
      "loss": 0.5757610201835632,
      "step": 196
    },
    {
      "epoch": 0.7888446215139442,
      "grad_norm": 0.6055193543434143,
      "learning_rate": 4.74439315111157e-06,
      "loss": 0.7602417469024658,
      "step": 198
    },
    {
      "epoch": 0.796812749003984,
      "grad_norm": 0.5638260841369629,
      "learning_rate": 4.737482633619234e-06,
      "loss": 1.0164211988449097,
      "step": 200
    },
    {
      "epoch": 0.8047808764940239,
      "grad_norm": 0.17684662342071533,
      "learning_rate": 4.730485723693912e-06,
      "loss": 0.9623653888702393,
      "step": 202
    },
    {
      "epoch": 0.8127490039840638,
      "grad_norm": 0.6719998717308044,
      "learning_rate": 4.723402725479177e-06,
      "loss": 1.4933321475982666,
      "step": 204
    },
    {
      "epoch": 0.8207171314741036,
      "grad_norm": 0.40897539258003235,
      "learning_rate": 4.71623394686071e-06,
      "loss": 1.2003211975097656,
      "step": 206
    },
    {
      "epoch": 0.8286852589641435,
      "grad_norm": 0.42933982610702515,
      "learning_rate": 4.708979699452923e-06,
      "loss": 0.6066120862960815,
      "step": 208
    },
    {
      "epoch": 0.8366533864541833,
      "grad_norm": 0.3967852294445038,
      "learning_rate": 4.701640298585405e-06,
      "loss": 1.2429534196853638,
      "step": 210
    },
    {
      "epoch": 0.8446215139442231,
      "grad_norm": 0.4494732618331909,
      "learning_rate": 4.694216063289219e-06,
      "loss": 1.1972289085388184,
      "step": 212
    },
    {
      "epoch": 0.852589641434263,
      "grad_norm": 0.464714378118515,
      "learning_rate": 4.686707316283033e-06,
      "loss": 0.9360241889953613,
      "step": 214
    },
    {
      "epoch": 0.8605577689243028,
      "grad_norm": 1.3721944093704224,
      "learning_rate": 4.679114383959095e-06,
      "loss": 1.4072277545928955,
      "step": 216
    },
    {
      "epoch": 0.8685258964143426,
      "grad_norm": 1.063977837562561,
      "learning_rate": 4.6714375963690415e-06,
      "loss": 0.8383471965789795,
      "step": 218
    },
    {
      "epoch": 0.8764940239043825,
      "grad_norm": 0.37765106558799744,
      "learning_rate": 4.6636772872095545e-06,
      "loss": 1.2202045917510986,
      "step": 220
    },
    {
      "epoch": 0.8844621513944223,
      "grad_norm": 0.503717303276062,
      "learning_rate": 4.6558337938078505e-06,
      "loss": 1.4709622859954834,
      "step": 222
    },
    {
      "epoch": 0.8924302788844621,
      "grad_norm": 3.107658863067627,
      "learning_rate": 4.647907457107025e-06,
      "loss": 0.6914993524551392,
      "step": 224
    },
    {
      "epoch": 0.900398406374502,
      "grad_norm": 0.8120009899139404,
      "learning_rate": 4.639898621651226e-06,
      "loss": 0.9661911129951477,
      "step": 226
    },
    {
      "epoch": 0.9083665338645418,
      "grad_norm": 0.8294084668159485,
      "learning_rate": 4.631807635570681e-06,
      "loss": 1.0741502046585083,
      "step": 228
    },
    {
      "epoch": 0.9163346613545816,
      "grad_norm": 2.316786527633667,
      "learning_rate": 4.623634850566563e-06,
      "loss": 1.1733441352844238,
      "step": 230
    },
    {
      "epoch": 0.9243027888446215,
      "grad_norm": 0.6197091937065125,
      "learning_rate": 4.615380621895701e-06,
      "loss": 1.3102484941482544,
      "step": 232
    },
    {
      "epoch": 0.9322709163346613,
      "grad_norm": 1.104339599609375,
      "learning_rate": 4.607045308355141e-06,
      "loss": 0.802425742149353,
      "step": 234
    },
    {
      "epoch": 0.9402390438247012,
      "grad_norm": 0.3975062370300293,
      "learning_rate": 4.598629272266549e-06,
      "loss": 0.8443046808242798,
      "step": 236
    },
    {
      "epoch": 0.9482071713147411,
      "grad_norm": 0.5752636194229126,
      "learning_rate": 4.590132879460458e-06,
      "loss": 0.44599035382270813,
      "step": 238
    },
    {
      "epoch": 0.9561752988047809,
      "grad_norm": 3.8548107147216797,
      "learning_rate": 4.5815564992603705e-06,
      "loss": 0.8519377708435059,
      "step": 240
    },
    {
      "epoch": 0.9641434262948207,
      "grad_norm": 0.612487256526947,
      "learning_rate": 4.572900504466702e-06,
      "loss": 1.1212224960327148,
      "step": 242
    },
    {
      "epoch": 0.9721115537848606,
      "grad_norm": 0.4580307900905609,
      "learning_rate": 4.564165271340577e-06,
      "loss": 1.1981385946273804,
      "step": 244
    },
    {
      "epoch": 0.9800796812749004,
      "grad_norm": 1.678839087486267,
      "learning_rate": 4.555351179587472e-06,
      "loss": 0.9461389183998108,
      "step": 246
    },
    {
      "epoch": 0.9880478087649402,
      "grad_norm": 0.4141462445259094,
      "learning_rate": 4.546458612340714e-06,
      "loss": 0.9613786935806274,
      "step": 248
    },
    {
      "epoch": 0.9960159362549801,
      "grad_norm": 0.6734077334403992,
      "learning_rate": 4.537487956144821e-06,
      "loss": 0.9624918103218079,
      "step": 250
    },
    {
      "epoch": 1.00398406374502,
      "grad_norm": 0.7532833814620972,
      "learning_rate": 4.528439600938706e-06,
      "loss": 0.8902125954627991,
      "step": 252
    },
    {
      "epoch": 1.0119521912350598,
      "grad_norm": 0.6221255660057068,
      "learning_rate": 4.519313940038721e-06,
      "loss": 0.6846274137496948,
      "step": 254
    },
    {
      "epoch": 1.0199203187250996,
      "grad_norm": 0.4263109862804413,
      "learning_rate": 4.510111370121563e-06,
      "loss": 1.1590865850448608,
      "step": 256
    },
    {
      "epoch": 1.0278884462151394,
      "grad_norm": 0.4521820545196533,
      "learning_rate": 4.500832291207033e-06,
      "loss": 1.1081089973449707,
      "step": 258
    },
    {
      "epoch": 1.0358565737051793,
      "grad_norm": 0.796542227268219,
      "learning_rate": 4.491477106640646e-06,
      "loss": 1.0672179460525513,
      "step": 260
    },
    {
      "epoch": 1.043824701195219,
      "grad_norm": 0.4541199505329132,
      "learning_rate": 4.482046223076092e-06,
      "loss": 0.9890964031219482,
      "step": 262
    },
    {
      "epoch": 1.051792828685259,
      "grad_norm": 0.700903058052063,
      "learning_rate": 4.4725400504575725e-06,
      "loss": 1.0773767232894897,
      "step": 264
    },
    {
      "epoch": 1.0597609561752988,
      "grad_norm": 1.8440330028533936,
      "learning_rate": 4.462959002001971e-06,
      "loss": 0.6484236121177673,
      "step": 266
    },
    {
      "epoch": 1.0677290836653386,
      "grad_norm": 0.25330546498298645,
      "learning_rate": 4.453303494180895e-06,
      "loss": 0.9161630272865295,
      "step": 268
    },
    {
      "epoch": 1.0756972111553784,
      "grad_norm": 0.520406186580658,
      "learning_rate": 4.443573946702573e-06,
      "loss": 0.5697556138038635,
      "step": 270
    },
    {
      "epoch": 1.0836653386454183,
      "grad_norm": 0.4949530065059662,
      "learning_rate": 4.433770782493606e-06,
      "loss": 1.2091208696365356,
      "step": 272
    },
    {
      "epoch": 1.091633466135458,
      "grad_norm": 0.8702555894851685,
      "learning_rate": 4.42389442768059e-06,
      "loss": 0.645495593547821,
      "step": 274
    },
    {
      "epoch": 1.099601593625498,
      "grad_norm": 0.4450341761112213,
      "learning_rate": 4.413945311571586e-06,
      "loss": 0.8727758526802063,
      "step": 276
    },
    {
      "epoch": 1.1075697211155378,
      "grad_norm": 0.45327526330947876,
      "learning_rate": 4.40392386663747e-06,
      "loss": 0.8436755537986755,
      "step": 278
    },
    {
      "epoch": 1.1155378486055776,
      "grad_norm": 0.3597104251384735,
      "learning_rate": 4.393830528493118e-06,
      "loss": 0.989022970199585,
      "step": 280
    },
    {
      "epoch": 1.1235059760956174,
      "grad_norm": 0.44877246022224426,
      "learning_rate": 4.3836657358784875e-06,
      "loss": 1.1293596029281616,
      "step": 282
    },
    {
      "epoch": 1.1314741035856573,
      "grad_norm": 0.6172301173210144,
      "learning_rate": 4.373429930639534e-06,
      "loss": 1.1756088733673096,
      "step": 284
    },
    {
      "epoch": 1.139442231075697,
      "grad_norm": 1.675234317779541,
      "learning_rate": 4.3631235577090085e-06,
      "loss": 0.2694728374481201,
      "step": 286
    },
    {
      "epoch": 1.1474103585657371,
      "grad_norm": 0.17762239277362823,
      "learning_rate": 4.35274706508712e-06,
      "loss": 0.6856278777122498,
      "step": 288
    },
    {
      "epoch": 1.155378486055777,
      "grad_norm": 0.4809702932834625,
      "learning_rate": 4.342300903822057e-06,
      "loss": 0.8659718036651611,
      "step": 290
    },
    {
      "epoch": 1.1633466135458168,
      "grad_norm": 0.4551142752170563,
      "learning_rate": 4.331785527990384e-06,
      "loss": 0.44279271364212036,
      "step": 292
    },
    {
      "epoch": 1.1713147410358566,
      "grad_norm": 0.5981038212776184,
      "learning_rate": 4.3212013946773045e-06,
      "loss": 1.1593934297561646,
      "step": 294
    },
    {
      "epoch": 1.1792828685258965,
      "grad_norm": 0.38801270723342896,
      "learning_rate": 4.31054896395679e-06,
      "loss": 1.199198603630066,
      "step": 296
    },
    {
      "epoch": 1.1872509960159363,
      "grad_norm": 0.3859914541244507,
      "learning_rate": 4.299828698871581e-06,
      "loss": 1.3288164138793945,
      "step": 298
    },
    {
      "epoch": 1.1952191235059761,
      "grad_norm": 1.0479567050933838,
      "learning_rate": 4.289041065413062e-06,
      "loss": 0.6392731070518494,
      "step": 300
    },
    {
      "epoch": 1.203187250996016,
      "grad_norm": 0.5702760219573975,
      "learning_rate": 4.278186532501003e-06,
      "loss": 0.9253982901573181,
      "step": 302
    },
    {
      "epoch": 1.2111553784860558,
      "grad_norm": 0.5659236311912537,
      "learning_rate": 4.267265571963179e-06,
      "loss": 0.5590313673019409,
      "step": 304
    },
    {
      "epoch": 1.2191235059760956,
      "grad_norm": 0.4407517910003662,
      "learning_rate": 4.256278658514858e-06,
      "loss": 1.1623272895812988,
      "step": 306
    },
    {
      "epoch": 1.2270916334661355,
      "grad_norm": 2.926598072052002,
      "learning_rate": 4.245226269738165e-06,
      "loss": 1.0200086832046509,
      "step": 308
    },
    {
      "epoch": 1.2350597609561753,
      "grad_norm": 0.600170373916626,
      "learning_rate": 4.23410888606133e-06,
      "loss": 1.1409556865692139,
      "step": 310
    },
    {
      "epoch": 1.2430278884462151,
      "grad_norm": 1.3301829099655151,
      "learning_rate": 4.222926990737793e-06,
      "loss": 1.0460419654846191,
      "step": 312
    },
    {
      "epoch": 1.250996015936255,
      "grad_norm": 1.3327184915542603,
      "learning_rate": 4.211681069825207e-06,
      "loss": 0.8847169280052185,
      "step": 314
    },
    {
      "epoch": 1.2589641434262948,
      "grad_norm": 0.9248475432395935,
      "learning_rate": 4.200371612164305e-06,
      "loss": 0.7445775866508484,
      "step": 316
    },
    {
      "epoch": 1.2669322709163346,
      "grad_norm": 1.0575453042984009,
      "learning_rate": 4.1889991093576555e-06,
      "loss": 0.6644352078437805,
      "step": 318
    },
    {
      "epoch": 1.2749003984063745,
      "grad_norm": 1.7681677341461182,
      "learning_rate": 4.177564055748284e-06,
      "loss": 0.6071811318397522,
      "step": 320
    },
    {
      "epoch": 1.2828685258964143,
      "grad_norm": 0.6115918159484863,
      "learning_rate": 4.166066948398199e-06,
      "loss": 0.9597835540771484,
      "step": 322
    },
    {
      "epoch": 1.2908366533864541,
      "grad_norm": 0.4188759922981262,
      "learning_rate": 4.154508287066772e-06,
      "loss": 0.8909115791320801,
      "step": 324
    },
    {
      "epoch": 1.298804780876494,
      "grad_norm": 0.5595813393592834,
      "learning_rate": 4.142888574189025e-06,
      "loss": 1.3452808856964111,
      "step": 326
    },
    {
      "epoch": 1.3067729083665338,
      "grad_norm": 0.311170369386673,
      "learning_rate": 4.131208314853778e-06,
      "loss": 0.6613047122955322,
      "step": 328
    },
    {
      "epoch": 1.3147410358565736,
      "grad_norm": 0.5032985210418701,
      "learning_rate": 4.119468016781707e-06,
      "loss": 0.84752357006073,
      "step": 330
    },
    {
      "epoch": 1.3227091633466135,
      "grad_norm": 0.6040851473808289,
      "learning_rate": 4.107668190303266e-06,
      "loss": 0.9215185642242432,
      "step": 332
    },
    {
      "epoch": 1.3306772908366533,
      "grad_norm": 1.0362443923950195,
      "learning_rate": 4.095809348336506e-06,
      "loss": 0.6224943399429321,
      "step": 334
    },
    {
      "epoch": 1.3386454183266931,
      "grad_norm": 3.144315242767334,
      "learning_rate": 4.083892006364777e-06,
      "loss": 0.8689337372779846,
      "step": 336
    },
    {
      "epoch": 1.3466135458167332,
      "grad_norm": 1.416516661643982,
      "learning_rate": 4.071916682414327e-06,
      "loss": 1.1079444885253906,
      "step": 338
    },
    {
      "epoch": 1.354581673306773,
      "grad_norm": 0.6034834384918213,
      "learning_rate": 4.059883897031777e-06,
      "loss": 1.0986981391906738,
      "step": 340
    },
    {
      "epoch": 1.3625498007968129,
      "grad_norm": 0.41596922278404236,
      "learning_rate": 4.0477941732615e-06,
      "loss": 1.1446994543075562,
      "step": 342
    },
    {
      "epoch": 1.3705179282868527,
      "grad_norm": 0.6623582243919373,
      "learning_rate": 4.035648036622879e-06,
      "loss": 1.1056511402130127,
      "step": 344
    },
    {
      "epoch": 1.3784860557768925,
      "grad_norm": 0.6701726317405701,
      "learning_rate": 4.023446015087469e-06,
      "loss": 0.6848117709159851,
      "step": 346
    },
    {
      "epoch": 1.3864541832669324,
      "grad_norm": 0.5770328640937805,
      "learning_rate": 4.0111886390560446e-06,
      "loss": 1.1336954832077026,
      "step": 348
    },
    {
      "epoch": 1.3944223107569722,
      "grad_norm": 0.5340338349342346,
      "learning_rate": 3.998876441335544e-06,
      "loss": 0.6723015308380127,
      "step": 350
    },
    {
      "epoch": 1.402390438247012,
      "grad_norm": 5.6478495597839355,
      "learning_rate": 3.98650995711591e-06,
      "loss": 1.1838804483413696,
      "step": 352
    },
    {
      "epoch": 1.4103585657370519,
      "grad_norm": 1.6889629364013672,
      "learning_rate": 3.974089723946826e-06,
      "loss": 1.0962597131729126,
      "step": 354
    },
    {
      "epoch": 1.4183266932270917,
      "grad_norm": 1.274549961090088,
      "learning_rate": 3.9616162817143465e-06,
      "loss": 0.9171859622001648,
      "step": 356
    },
    {
      "epoch": 1.4262948207171315,
      "grad_norm": 0.9090215563774109,
      "learning_rate": 3.949090172617437e-06,
      "loss": 1.1338566541671753,
      "step": 358
    },
    {
      "epoch": 1.4342629482071714,
      "grad_norm": 0.8088547587394714,
      "learning_rate": 3.936511941144397e-06,
      "loss": 0.5907384753227234,
      "step": 360
    },
    {
      "epoch": 1.4422310756972112,
      "grad_norm": 0.31662285327911377,
      "learning_rate": 3.9238821340491936e-06,
      "loss": 0.7013169527053833,
      "step": 362
    },
    {
      "epoch": 1.450199203187251,
      "grad_norm": 2.500298023223877,
      "learning_rate": 3.911201300327703e-06,
      "loss": 1.0219976902008057,
      "step": 364
    },
    {
      "epoch": 1.4581673306772909,
      "grad_norm": 0.46223050355911255,
      "learning_rate": 3.898469991193835e-06,
      "loss": 1.1217302083969116,
      "step": 366
    },
    {
      "epoch": 1.4661354581673307,
      "grad_norm": 0.4489881098270416,
      "learning_rate": 3.885688760055581e-06,
      "loss": 0.5916672348976135,
      "step": 368
    },
    {
      "epoch": 1.4741035856573705,
      "grad_norm": 0.3636326193809509,
      "learning_rate": 3.872858162490956e-06,
      "loss": 0.6642428040504456,
      "step": 370
    },
    {
      "epoch": 1.4820717131474104,
      "grad_norm": 0.7223714590072632,
      "learning_rate": 3.859978756223845e-06,
      "loss": 1.2154208421707153,
      "step": 372
    },
    {
      "epoch": 1.4900398406374502,
      "grad_norm": 0.3954339325428009,
      "learning_rate": 3.847051101099768e-06,
      "loss": 1.1045606136322021,
      "step": 374
    },
    {
      "epoch": 1.49800796812749,
      "grad_norm": 0.6890347003936768,
      "learning_rate": 3.834075759061535e-06,
      "loss": 1.1236262321472168,
      "step": 376
    },
    {
      "epoch": 1.5059760956175299,
      "grad_norm": 4.405030250549316,
      "learning_rate": 3.8210532941248255e-06,
      "loss": 0.9026872515678406,
      "step": 378
    },
    {
      "epoch": 1.5139442231075697,
      "grad_norm": 0.40552160143852234,
      "learning_rate": 3.8079842723536688e-06,
      "loss": 1.105115294456482,
      "step": 380
    },
    {
      "epoch": 1.5219123505976095,
      "grad_norm": 0.8462726473808289,
      "learning_rate": 3.7948692618358424e-06,
      "loss": 0.7042898535728455,
      "step": 382
    },
    {
      "epoch": 1.5298804780876494,
      "grad_norm": 0.6684491038322449,
      "learning_rate": 3.781708832658169e-06,
      "loss": 0.7331148386001587,
      "step": 384
    },
    {
      "epoch": 1.5378486055776892,
      "grad_norm": 0.8122864365577698,
      "learning_rate": 3.7685035568817492e-06,
      "loss": 1.0483731031417847,
      "step": 386
    },
    {
      "epoch": 1.545816733067729,
      "grad_norm": 1.0633962154388428,
      "learning_rate": 3.7552540085170826e-06,
      "loss": 1.1307730674743652,
      "step": 388
    },
    {
      "epoch": 1.5537848605577689,
      "grad_norm": 2.401926279067993,
      "learning_rate": 3.7419607634991235e-06,
      "loss": 0.5506023168563843,
      "step": 390
    },
    {
      "epoch": 1.5617529880478087,
      "grad_norm": 0.6989754438400269,
      "learning_rate": 3.7286243996622432e-06,
      "loss": 0.5932711362838745,
      "step": 392
    },
    {
      "epoch": 1.5697211155378485,
      "grad_norm": 10.104994773864746,
      "learning_rate": 3.715245496715116e-06,
      "loss": 1.1775317192077637,
      "step": 394
    },
    {
      "epoch": 1.5776892430278884,
      "grad_norm": 0.6455057859420776,
      "learning_rate": 3.7018246362155134e-06,
      "loss": 1.0059901475906372,
      "step": 396
    },
    {
      "epoch": 1.5856573705179282,
      "grad_norm": 0.2941265106201172,
      "learning_rate": 3.6883624015450344e-06,
      "loss": 1.1076565980911255,
      "step": 398
    },
    {
      "epoch": 1.593625498007968,
      "grad_norm": 0.573162317276001,
      "learning_rate": 3.6748593778837367e-06,
      "loss": 0.5877851247787476,
      "step": 400
    },
    {
      "epoch": 1.6015936254980079,
      "grad_norm": 0.43955594301223755,
      "learning_rate": 3.661316152184708e-06,
      "loss": 1.089148759841919,
      "step": 402
    },
    {
      "epoch": 1.6095617529880477,
      "grad_norm": 0.5078707933425903,
      "learning_rate": 3.6477333131485483e-06,
      "loss": 1.2552837133407593,
      "step": 404
    },
    {
      "epoch": 1.6175298804780875,
      "grad_norm": 0.8013210892677307,
      "learning_rate": 3.634111451197781e-06,
      "loss": 0.7477819919586182,
      "step": 406
    },
    {
      "epoch": 1.6254980079681274,
      "grad_norm": 0.7095657587051392,
      "learning_rate": 3.6204511584511865e-06,
      "loss": 0.5438832640647888,
      "step": 408
    },
    {
      "epoch": 1.6334661354581672,
      "grad_norm": 0.4615778625011444,
      "learning_rate": 3.606753028698067e-06,
      "loss": 0.7296803593635559,
      "step": 410
    },
    {
      "epoch": 1.641434262948207,
      "grad_norm": 1.9634928703308105,
      "learning_rate": 3.593017657372436e-06,
      "loss": 0.49200546741485596,
      "step": 412
    },
    {
      "epoch": 1.6494023904382469,
      "grad_norm": 0.29043668508529663,
      "learning_rate": 3.5792456415271288e-06,
      "loss": 0.8374850153923035,
      "step": 414
    },
    {
      "epoch": 1.6573705179282867,
      "grad_norm": 0.5617873072624207,
      "learning_rate": 3.565437579807858e-06,
      "loss": 0.7375838756561279,
      "step": 416
    },
    {
      "epoch": 1.6653386454183265,
      "grad_norm": 1.787645697593689,
      "learning_rate": 3.551594072427188e-06,
      "loss": 0.9972248077392578,
      "step": 418
    },
    {
      "epoch": 1.6733067729083664,
      "grad_norm": 0.3820089101791382,
      "learning_rate": 3.537715721138444e-06,
      "loss": 1.1183611154556274,
      "step": 420
    },
    {
      "epoch": 1.6812749003984062,
      "grad_norm": 0.4002365469932556,
      "learning_rate": 3.523803129209554e-06,
      "loss": 0.9985948801040649,
      "step": 422
    },
    {
      "epoch": 1.6892430278884463,
      "grad_norm": 0.617779552936554,
      "learning_rate": 3.5098569013968314e-06,
      "loss": 0.6952139139175415,
      "step": 424
    },
    {
      "epoch": 1.697211155378486,
      "grad_norm": 0.4030837416648865,
      "learning_rate": 3.49587764391868e-06,
      "loss": 1.1934587955474854,
      "step": 426
    },
    {
      "epoch": 1.705179282868526,
      "grad_norm": 0.3113058805465698,
      "learning_rate": 3.4818659644292473e-06,
      "loss": 1.1069426536560059,
      "step": 428
    },
    {
      "epoch": 1.7131474103585658,
      "grad_norm": 0.43936672806739807,
      "learning_rate": 3.4678224719920085e-06,
      "loss": 0.8586387038230896,
      "step": 430
    },
    {
      "epoch": 1.7211155378486056,
      "grad_norm": 0.7433982491493225,
      "learning_rate": 3.4537477770532933e-06,
      "loss": 0.950721263885498,
      "step": 432
    },
    {
      "epoch": 1.7290836653386454,
      "grad_norm": 1.082598090171814,
      "learning_rate": 3.4396424914157527e-06,
      "loss": 0.6536106467247009,
      "step": 434
    },
    {
      "epoch": 1.7370517928286853,
      "grad_norm": 1.1421170234680176,
      "learning_rate": 3.4255072282117574e-06,
      "loss": 0.40250933170318604,
      "step": 436
    },
    {
      "epoch": 1.745019920318725,
      "grad_norm": 0.6019651889801025,
      "learning_rate": 3.411342601876754e-06,
      "loss": 0.7538256645202637,
      "step": 438
    },
    {
      "epoch": 1.752988047808765,
      "grad_norm": 0.7230652570724487,
      "learning_rate": 3.3971492281225554e-06,
      "loss": 0.6363106369972229,
      "step": 440
    },
    {
      "epoch": 1.7609561752988048,
      "grad_norm": 0.7830895185470581,
      "learning_rate": 3.382927723910572e-06,
      "loss": 1.5372300148010254,
      "step": 442
    },
    {
      "epoch": 1.7689243027888446,
      "grad_norm": 0.5875637531280518,
      "learning_rate": 3.3686787074249966e-06,
      "loss": 0.8283383250236511,
      "step": 444
    },
    {
      "epoch": 1.7768924302788844,
      "grad_norm": 0.506118893623352,
      "learning_rate": 3.354402798045935e-06,
      "loss": 1.0632597208023071,
      "step": 446
    },
    {
      "epoch": 1.7848605577689243,
      "grad_norm": 0.43126431107521057,
      "learning_rate": 3.340100616322479e-06,
      "loss": 1.0956040620803833,
      "step": 448
    },
    {
      "epoch": 1.792828685258964,
      "grad_norm": 0.5031911730766296,
      "learning_rate": 3.3257727839457356e-06,
      "loss": 0.8361770510673523,
      "step": 450
    },
    {
      "epoch": 1.800796812749004,
      "grad_norm": 0.2947663962841034,
      "learning_rate": 3.311419923721798e-06,
      "loss": 0.48412880301475525,
      "step": 452
    },
    {
      "epoch": 1.8087649402390438,
      "grad_norm": 0.5195736289024353,
      "learning_rate": 3.29704265954468e-06,
      "loss": 0.9252212643623352,
      "step": 454
    },
    {
      "epoch": 1.8167330677290838,
      "grad_norm": 0.8320481777191162,
      "learning_rate": 3.282641616369193e-06,
      "loss": 0.685675323009491,
      "step": 456
    },
    {
      "epoch": 1.8247011952191237,
      "grad_norm": 0.5695275664329529,
      "learning_rate": 3.2682174201837773e-06,
      "loss": 1.034195899963379,
      "step": 458
    },
    {
      "epoch": 1.8326693227091635,
      "grad_norm": 1.0242846012115479,
      "learning_rate": 3.2537706979832955e-06,
      "loss": 0.720179557800293,
      "step": 460
    },
    {
      "epoch": 1.8406374501992033,
      "grad_norm": 0.4475300908088684,
      "learning_rate": 3.23930207774178e-06,
      "loss": 0.7963992953300476,
      "step": 462
    },
    {
      "epoch": 1.8486055776892432,
      "grad_norm": 0.8903407454490662,
      "learning_rate": 3.2248121883851296e-06,
      "loss": 0.10920298844575882,
      "step": 464
    },
    {
      "epoch": 1.856573705179283,
      "grad_norm": 0.47660109400749207,
      "learning_rate": 3.210301659763777e-06,
      "loss": 1.129386067390442,
      "step": 466
    },
    {
      "epoch": 1.8645418326693228,
      "grad_norm": 0.5083760619163513,
      "learning_rate": 3.1957711226253073e-06,
      "loss": 0.5216770768165588,
      "step": 468
    },
    {
      "epoch": 1.8725099601593627,
      "grad_norm": 0.6058355569839478,
      "learning_rate": 3.181221208587041e-06,
      "loss": 1.1222081184387207,
      "step": 470
    },
    {
      "epoch": 1.8804780876494025,
      "grad_norm": 0.7565557360649109,
      "learning_rate": 3.16665255010858e-06,
      "loss": 1.0339274406433105,
      "step": 472
    },
    {
      "epoch": 1.8884462151394423,
      "grad_norm": 12.444122314453125,
      "learning_rate": 3.152065780464311e-06,
      "loss": 1.132270097732544,
      "step": 474
    },
    {
      "epoch": 1.8964143426294822,
      "grad_norm": 0.35113006830215454,
      "learning_rate": 3.1374615337158874e-06,
      "loss": 0.9474564790725708,
      "step": 476
    },
    {
      "epoch": 1.904382470119522,
      "grad_norm": 0.8079137206077576,
      "learning_rate": 3.1228404446846577e-06,
      "loss": 0.862366259098053,
      "step": 478
    },
    {
      "epoch": 1.9123505976095618,
      "grad_norm": 1.731661319732666,
      "learning_rate": 3.1082031489240775e-06,
      "loss": 0.604378879070282,
      "step": 480
    },
    {
      "epoch": 1.9203187250996017,
      "grad_norm": 0.6579747796058655,
      "learning_rate": 3.093550282692081e-06,
      "loss": 0.7587324976921082,
      "step": 482
    },
    {
      "epoch": 1.9282868525896415,
      "grad_norm": 0.49387723207473755,
      "learning_rate": 3.0788824829234206e-06,
      "loss": 1.118217945098877,
      "step": 484
    },
    {
      "epoch": 1.9362549800796813,
      "grad_norm": 0.38881438970565796,
      "learning_rate": 3.064200387201987e-06,
      "loss": 1.1538664102554321,
      "step": 486
    },
    {
      "epoch": 1.9442231075697212,
      "grad_norm": 0.3889444172382355,
      "learning_rate": 3.0495046337330917e-06,
      "loss": 1.0748200416564941,
      "step": 488
    },
    {
      "epoch": 1.952191235059761,
      "grad_norm": 0.4628307819366455,
      "learning_rate": 3.034795861315723e-06,
      "loss": 1.0917750597000122,
      "step": 490
    },
    {
      "epoch": 1.9601593625498008,
      "grad_norm": 0.3515513241291046,
      "learning_rate": 3.020074709314781e-06,
      "loss": 0.7694330811500549,
      "step": 492
    },
    {
      "epoch": 1.9681274900398407,
      "grad_norm": 2.0103540420532227,
      "learning_rate": 3.0053418176332844e-06,
      "loss": 0.6312040686607361,
      "step": 494
    },
    {
      "epoch": 1.9760956175298805,
      "grad_norm": 0.5194686651229858,
      "learning_rate": 2.9905978266845565e-06,
      "loss": 1.059701681137085,
      "step": 496
    },
    {
      "epoch": 1.9840637450199203,
      "grad_norm": 1.5531368255615234,
      "learning_rate": 2.975843377364385e-06,
      "loss": 0.8427159786224365,
      "step": 498
    },
    {
      "epoch": 1.9920318725099602,
      "grad_norm": 0.4374231994152069,
      "learning_rate": 2.961079111023167e-06,
      "loss": 1.1491104364395142,
      "step": 500
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7201674580574036,
      "learning_rate": 2.9463056694380263e-06,
      "loss": 0.5880916118621826,
      "step": 502
    },
    {
      "epoch": 2.00796812749004,
      "grad_norm": 0.3835521340370178,
      "learning_rate": 2.9315236947849213e-06,
      "loss": 0.6079206466674805,
      "step": 504
    },
    {
      "epoch": 2.0159362549800797,
      "grad_norm": 0.6948376297950745,
      "learning_rate": 2.916733829610726e-06,
      "loss": 0.5914795994758606,
      "step": 506
    },
    {
      "epoch": 2.0239043824701195,
      "grad_norm": 0.12258737534284592,
      "learning_rate": 2.901936716805302e-06,
      "loss": 0.5301275849342346,
      "step": 508
    },
    {
      "epoch": 2.0318725099601593,
      "grad_norm": 0.388775497674942,
      "learning_rate": 2.887132999573554e-06,
      "loss": 0.8716145753860474,
      "step": 510
    },
    {
      "epoch": 2.039840637450199,
      "grad_norm": 0.5945245623588562,
      "learning_rate": 2.8723233214074675e-06,
      "loss": 0.7388210892677307,
      "step": 512
    },
    {
      "epoch": 2.047808764940239,
      "grad_norm": 5.118804931640625,
      "learning_rate": 2.857508326058142e-06,
      "loss": 0.3199237287044525,
      "step": 514
    },
    {
      "epoch": 2.055776892430279,
      "grad_norm": 0.8427819013595581,
      "learning_rate": 2.842688657507805e-06,
      "loss": 0.8688737750053406,
      "step": 516
    },
    {
      "epoch": 2.0637450199203187,
      "grad_norm": 0.719185471534729,
      "learning_rate": 2.8278649599418194e-06,
      "loss": 0.5541288256645203,
      "step": 518
    },
    {
      "epoch": 2.0717131474103585,
      "grad_norm": 0.9245908856391907,
      "learning_rate": 2.813037877720683e-06,
      "loss": 0.8925630450248718,
      "step": 520
    },
    {
      "epoch": 2.0796812749003983,
      "grad_norm": 0.5760779976844788,
      "learning_rate": 2.798208055352017e-06,
      "loss": 0.9683179259300232,
      "step": 522
    },
    {
      "epoch": 2.087649402390438,
      "grad_norm": 0.9741346836090088,
      "learning_rate": 2.7833761374625546e-06,
      "loss": 0.9222256541252136,
      "step": 524
    },
    {
      "epoch": 2.095617529880478,
      "grad_norm": 1.7210060358047485,
      "learning_rate": 2.7685427687701172e-06,
      "loss": 0.4816100299358368,
      "step": 526
    },
    {
      "epoch": 2.103585657370518,
      "grad_norm": 0.5505311489105225,
      "learning_rate": 2.753708594055588e-06,
      "loss": 0.7380607724189758,
      "step": 528
    },
    {
      "epoch": 2.1115537848605577,
      "grad_norm": 0.20451760292053223,
      "learning_rate": 2.7388742581348893e-06,
      "loss": 0.6325430274009705,
      "step": 530
    },
    {
      "epoch": 2.1195219123505975,
      "grad_norm": 0.63947993516922,
      "learning_rate": 2.7240404058309487e-06,
      "loss": 0.24245773255825043,
      "step": 532
    },
    {
      "epoch": 2.1274900398406373,
      "grad_norm": 0.5770893692970276,
      "learning_rate": 2.709207681945673e-06,
      "loss": 0.8044096827507019,
      "step": 534
    },
    {
      "epoch": 2.135458167330677,
      "grad_norm": 2.8005356788635254,
      "learning_rate": 2.6943767312319167e-06,
      "loss": 0.674305260181427,
      "step": 536
    },
    {
      "epoch": 2.143426294820717,
      "grad_norm": 0.5809776186943054,
      "learning_rate": 2.679548198365462e-06,
      "loss": 0.6474828720092773,
      "step": 538
    },
    {
      "epoch": 2.151394422310757,
      "grad_norm": 0.5860468745231628,
      "learning_rate": 2.6647227279169873e-06,
      "loss": 0.9396933317184448,
      "step": 540
    },
    {
      "epoch": 2.1593625498007967,
      "grad_norm": 0.4746709167957306,
      "learning_rate": 2.6499009643240527e-06,
      "loss": 0.8665639162063599,
      "step": 542
    },
    {
      "epoch": 2.1673306772908365,
      "grad_norm": 1.7936333417892456,
      "learning_rate": 2.6350835518630907e-06,
      "loss": 0.8066052198410034,
      "step": 544
    },
    {
      "epoch": 2.1752988047808763,
      "grad_norm": 0.5707678198814392,
      "learning_rate": 2.620271134621397e-06,
      "loss": 0.6862360835075378,
      "step": 546
    },
    {
      "epoch": 2.183266932270916,
      "grad_norm": 0.5736294388771057,
      "learning_rate": 2.605464356469131e-06,
      "loss": 0.9289900660514832,
      "step": 548
    },
    {
      "epoch": 2.191235059760956,
      "grad_norm": 0.3922679126262665,
      "learning_rate": 2.5906638610313336e-06,
      "loss": 1.0125665664672852,
      "step": 550
    },
    {
      "epoch": 2.199203187250996,
      "grad_norm": 0.6215779781341553,
      "learning_rate": 2.575870291659946e-06,
      "loss": 0.7937463521957397,
      "step": 552
    },
    {
      "epoch": 2.2071713147410357,
      "grad_norm": 0.6370753049850464,
      "learning_rate": 2.5610842914058446e-06,
      "loss": 0.45048844814300537,
      "step": 554
    },
    {
      "epoch": 2.2151394422310755,
      "grad_norm": 0.48614996671676636,
      "learning_rate": 2.5463065029908906e-06,
      "loss": 0.6716344952583313,
      "step": 556
    },
    {
      "epoch": 2.2231075697211153,
      "grad_norm": 0.9526214003562927,
      "learning_rate": 2.5315375687799896e-06,
      "loss": 0.8591071367263794,
      "step": 558
    },
    {
      "epoch": 2.231075697211155,
      "grad_norm": 0.34850233793258667,
      "learning_rate": 2.5167781307531714e-06,
      "loss": 0.9045411944389343,
      "step": 560
    },
    {
      "epoch": 2.239043824701195,
      "grad_norm": 1.1935197114944458,
      "learning_rate": 2.5020288304776822e-06,
      "loss": 1.000231146812439,
      "step": 562
    },
    {
      "epoch": 2.247011952191235,
      "grad_norm": 1.1329901218414307,
      "learning_rate": 2.487290309080097e-06,
      "loss": 0.7996300458908081,
      "step": 564
    },
    {
      "epoch": 2.2549800796812747,
      "grad_norm": 0.5488888621330261,
      "learning_rate": 2.472563207218454e-06,
      "loss": 0.6275344491004944,
      "step": 566
    },
    {
      "epoch": 2.2629482071713145,
      "grad_norm": 0.3373243510723114,
      "learning_rate": 2.457848165054402e-06,
      "loss": 0.7087105512619019,
      "step": 568
    },
    {
      "epoch": 2.2709163346613543,
      "grad_norm": 0.4191913604736328,
      "learning_rate": 2.4431458222253755e-06,
      "loss": 0.9698747396469116,
      "step": 570
    },
    {
      "epoch": 2.278884462151394,
      "grad_norm": 0.6732890009880066,
      "learning_rate": 2.4284568178167923e-06,
      "loss": 0.6233561038970947,
      "step": 572
    },
    {
      "epoch": 2.2868525896414345,
      "grad_norm": 0.40283727645874023,
      "learning_rate": 2.413781790334272e-06,
      "loss": 0.7267651557922363,
      "step": 574
    },
    {
      "epoch": 2.2948207171314743,
      "grad_norm": 1.7338671684265137,
      "learning_rate": 2.3991213776758776e-06,
      "loss": 0.32845792174339294,
      "step": 576
    },
    {
      "epoch": 2.302788844621514,
      "grad_norm": 0.5662968158721924,
      "learning_rate": 2.3844762171043975e-06,
      "loss": 0.6570364832878113,
      "step": 578
    },
    {
      "epoch": 2.310756972111554,
      "grad_norm": 0.6602838039398193,
      "learning_rate": 2.3698469452196328e-06,
      "loss": 0.6247450709342957,
      "step": 580
    },
    {
      "epoch": 2.318725099601594,
      "grad_norm": 1.3522082567214966,
      "learning_rate": 2.355234197930735e-06,
      "loss": 0.6848940849304199,
      "step": 582
    },
    {
      "epoch": 2.3266932270916336,
      "grad_norm": 1.3658299446105957,
      "learning_rate": 2.340638610428554e-06,
      "loss": 0.779475748538971,
      "step": 584
    },
    {
      "epoch": 2.3346613545816735,
      "grad_norm": 0.9785452485084534,
      "learning_rate": 2.3260608171580412e-06,
      "loss": 0.626722514629364,
      "step": 586
    },
    {
      "epoch": 2.3426294820717133,
      "grad_norm": 1.4936383962631226,
      "learning_rate": 2.311501451790658e-06,
      "loss": 0.8517850637435913,
      "step": 588
    },
    {
      "epoch": 2.350597609561753,
      "grad_norm": 0.45495304465293884,
      "learning_rate": 2.2969611471968386e-06,
      "loss": 1.0095319747924805,
      "step": 590
    },
    {
      "epoch": 2.358565737051793,
      "grad_norm": 0.43414270877838135,
      "learning_rate": 2.2824405354184776e-06,
      "loss": 0.9951021671295166,
      "step": 592
    },
    {
      "epoch": 2.366533864541833,
      "grad_norm": 0.6097609996795654,
      "learning_rate": 2.267940247641458e-06,
      "loss": 0.6759450435638428,
      "step": 594
    },
    {
      "epoch": 2.3745019920318726,
      "grad_norm": 0.29373353719711304,
      "learning_rate": 2.2534609141682135e-06,
      "loss": 0.9598254561424255,
      "step": 596
    },
    {
      "epoch": 2.3824701195219125,
      "grad_norm": 0.29972609877586365,
      "learning_rate": 2.2390031643903286e-06,
      "loss": 0.7291547656059265,
      "step": 598
    },
    {
      "epoch": 2.3904382470119523,
      "grad_norm": 1.2946021556854248,
      "learning_rate": 2.224567626761185e-06,
      "loss": 0.9443849325180054,
      "step": 600
    },
    {
      "epoch": 2.398406374501992,
      "grad_norm": 0.38664981722831726,
      "learning_rate": 2.2101549287686384e-06,
      "loss": 0.9774645566940308,
      "step": 602
    },
    {
      "epoch": 2.406374501992032,
      "grad_norm": 0.6325794458389282,
      "learning_rate": 2.195765696907745e-06,
      "loss": 0.5516265034675598,
      "step": 604
    },
    {
      "epoch": 2.414342629482072,
      "grad_norm": 0.5667062401771545,
      "learning_rate": 2.1814005566535303e-06,
      "loss": 0.23102697730064392,
      "step": 606
    },
    {
      "epoch": 2.4223107569721116,
      "grad_norm": 0.2469281256198883,
      "learning_rate": 2.167060132433798e-06,
      "loss": 0.5323957204818726,
      "step": 608
    },
    {
      "epoch": 2.4302788844621515,
      "grad_norm": 0.1867014318704605,
      "learning_rate": 2.1527450476019895e-06,
      "loss": 0.8783145546913147,
      "step": 610
    },
    {
      "epoch": 2.4382470119521913,
      "grad_norm": 0.3423818349838257,
      "learning_rate": 2.138455924410085e-06,
      "loss": 0.7297731637954712,
      "step": 612
    },
    {
      "epoch": 2.446215139442231,
      "grad_norm": 0.48558947443962097,
      "learning_rate": 2.1241933839815606e-06,
      "loss": 0.5452163219451904,
      "step": 614
    },
    {
      "epoch": 2.454183266932271,
      "grad_norm": 0.4644261300563812,
      "learning_rate": 2.109958046284382e-06,
      "loss": 0.8939840197563171,
      "step": 616
    },
    {
      "epoch": 2.462151394422311,
      "grad_norm": 1.2982784509658813,
      "learning_rate": 2.0957505301040604e-06,
      "loss": 0.7693812251091003,
      "step": 618
    },
    {
      "epoch": 2.4701195219123506,
      "grad_norm": 1.4525824785232544,
      "learning_rate": 2.081571453016756e-06,
      "loss": 0.7849122285842896,
      "step": 620
    },
    {
      "epoch": 2.4780876494023905,
      "grad_norm": 0.5957829356193542,
      "learning_rate": 2.067421431362427e-06,
      "loss": 0.8612769246101379,
      "step": 622
    },
    {
      "epoch": 2.4860557768924303,
      "grad_norm": 0.7306169271469116,
      "learning_rate": 2.053301080218044e-06,
      "loss": 0.7797048091888428,
      "step": 624
    },
    {
      "epoch": 2.49402390438247,
      "grad_norm": 0.5810465216636658,
      "learning_rate": 2.039211013370853e-06,
      "loss": 0.5878288745880127,
      "step": 626
    },
    {
      "epoch": 2.50199203187251,
      "grad_norm": 0.6188574433326721,
      "learning_rate": 2.025151843291693e-06,
      "loss": 1.026731014251709,
      "step": 628
    },
    {
      "epoch": 2.50996015936255,
      "grad_norm": 0.4957115650177002,
      "learning_rate": 2.0111241811083733e-06,
      "loss": 0.48043641448020935,
      "step": 630
    },
    {
      "epoch": 2.5179282868525896,
      "grad_norm": 1.5852978229522705,
      "learning_rate": 1.9971286365791092e-06,
      "loss": 0.5416110754013062,
      "step": 632
    },
    {
      "epoch": 2.5258964143426295,
      "grad_norm": 0.3960880637168884,
      "learning_rate": 1.9831658180660197e-06,
      "loss": 0.9946979880332947,
      "step": 634
    },
    {
      "epoch": 2.5338645418326693,
      "grad_norm": 0.44946765899658203,
      "learning_rate": 1.9692363325086774e-06,
      "loss": 0.9242001175880432,
      "step": 636
    },
    {
      "epoch": 2.541832669322709,
      "grad_norm": 1.37661874294281,
      "learning_rate": 1.95534078539773e-06,
      "loss": 0.9803711175918579,
      "step": 638
    },
    {
      "epoch": 2.549800796812749,
      "grad_norm": 0.25824299454689026,
      "learning_rate": 1.9414797807485796e-06,
      "loss": 0.81007981300354,
      "step": 640
    },
    {
      "epoch": 2.557768924302789,
      "grad_norm": 1.0457597970962524,
      "learning_rate": 1.9276539210751295e-06,
      "loss": 0.9650387167930603,
      "step": 642
    },
    {
      "epoch": 2.5657370517928286,
      "grad_norm": 0.5793036818504333,
      "learning_rate": 1.9138638073635886e-06,
      "loss": 1.1386988162994385,
      "step": 644
    },
    {
      "epoch": 2.5737051792828685,
      "grad_norm": 0.4966047406196594,
      "learning_rate": 1.90011003904635e-06,
      "loss": 0.5513513684272766,
      "step": 646
    },
    {
      "epoch": 2.5816733067729083,
      "grad_norm": 1.1551893949508667,
      "learning_rate": 1.886393213975939e-06,
      "loss": 0.9931833744049072,
      "step": 648
    },
    {
      "epoch": 2.589641434262948,
      "grad_norm": 0.26254934072494507,
      "learning_rate": 1.8727139283990203e-06,
      "loss": 0.7399964928627014,
      "step": 650
    },
    {
      "epoch": 2.597609561752988,
      "grad_norm": 0.5739290118217468,
      "learning_rate": 1.8590727769304789e-06,
      "loss": 0.847949743270874,
      "step": 652
    },
    {
      "epoch": 2.605577689243028,
      "grad_norm": 0.3866256773471832,
      "learning_rate": 1.8454703525275814e-06,
      "loss": 0.6400333046913147,
      "step": 654
    },
    {
      "epoch": 2.6135458167330676,
      "grad_norm": 0.5888515114784241,
      "learning_rate": 1.8319072464641912e-06,
      "loss": 0.3374575972557068,
      "step": 656
    },
    {
      "epoch": 2.6215139442231075,
      "grad_norm": 0.841457724571228,
      "learning_rate": 1.8183840483050741e-06,
      "loss": 0.5087389349937439,
      "step": 658
    },
    {
      "epoch": 2.6294820717131473,
      "grad_norm": 0.37636950612068176,
      "learning_rate": 1.8049013458802667e-06,
      "loss": 0.8487051725387573,
      "step": 660
    },
    {
      "epoch": 2.637450199203187,
      "grad_norm": 2.6286380290985107,
      "learning_rate": 1.7914597252595267e-06,
      "loss": 1.0009267330169678,
      "step": 662
    },
    {
      "epoch": 2.645418326693227,
      "grad_norm": 1.0971342325210571,
      "learning_rate": 1.7780597707268582e-06,
      "loss": 0.6267133355140686,
      "step": 664
    },
    {
      "epoch": 2.653386454183267,
      "grad_norm": 1.0502574443817139,
      "learning_rate": 1.7647020647551105e-06,
      "loss": 0.9501173496246338,
      "step": 666
    },
    {
      "epoch": 2.6613545816733066,
      "grad_norm": 0.23098650574684143,
      "learning_rate": 1.7513871879806605e-06,
      "loss": 0.5892425179481506,
      "step": 668
    },
    {
      "epoch": 2.6693227091633465,
      "grad_norm": 1.3565614223480225,
      "learning_rate": 1.7381157191781766e-06,
      "loss": 0.7291454076766968,
      "step": 670
    },
    {
      "epoch": 2.6772908366533863,
      "grad_norm": 0.15605708956718445,
      "learning_rate": 1.7248882352354571e-06,
      "loss": 0.40334466099739075,
      "step": 672
    },
    {
      "epoch": 2.685258964143426,
      "grad_norm": 0.4292656183242798,
      "learning_rate": 1.7117053111283528e-06,
      "loss": 1.0303913354873657,
      "step": 674
    },
    {
      "epoch": 2.6932270916334664,
      "grad_norm": 1.2042747735977173,
      "learning_rate": 1.6985675198957752e-06,
      "loss": 0.7125272154808044,
      "step": 676
    },
    {
      "epoch": 2.7011952191235062,
      "grad_norm": 0.5820655822753906,
      "learning_rate": 1.6854754326147916e-06,
      "loss": 0.6398980021476746,
      "step": 678
    },
    {
      "epoch": 2.709163346613546,
      "grad_norm": 0.33729568123817444,
      "learning_rate": 1.6724296183757925e-06,
      "loss": 0.975061297416687,
      "step": 680
    },
    {
      "epoch": 2.717131474103586,
      "grad_norm": 0.42134153842926025,
      "learning_rate": 1.6594306442577629e-06,
      "loss": 1.0035920143127441,
      "step": 682
    },
    {
      "epoch": 2.7250996015936257,
      "grad_norm": 0.5072264075279236,
      "learning_rate": 1.6464790753036298e-06,
      "loss": 0.6735944747924805,
      "step": 684
    },
    {
      "epoch": 2.7330677290836656,
      "grad_norm": 0.6527407169342041,
      "learning_rate": 1.6335754744956935e-06,
      "loss": 0.5050901770591736,
      "step": 686
    },
    {
      "epoch": 2.7410358565737054,
      "grad_norm": 0.6374827027320862,
      "learning_rate": 1.6207204027311685e-06,
      "loss": 0.9826086759567261,
      "step": 688
    },
    {
      "epoch": 2.7490039840637452,
      "grad_norm": 0.4688166677951813,
      "learning_rate": 1.6079144187977925e-06,
      "loss": 0.592976987361908,
      "step": 690
    },
    {
      "epoch": 2.756972111553785,
      "grad_norm": 0.48488789796829224,
      "learning_rate": 1.595158079349542e-06,
      "loss": 1.0681207180023193,
      "step": 692
    },
    {
      "epoch": 2.764940239043825,
      "grad_norm": 2.4061834812164307,
      "learning_rate": 1.5824519388824311e-06,
      "loss": 0.8421837687492371,
      "step": 694
    },
    {
      "epoch": 2.7729083665338647,
      "grad_norm": 1.4498945474624634,
      "learning_rate": 1.5697965497104117e-06,
      "loss": 0.7515864372253418,
      "step": 696
    },
    {
      "epoch": 2.7808764940239046,
      "grad_norm": 0.42357689142227173,
      "learning_rate": 1.557192461941367e-06,
      "loss": 0.8500891327857971,
      "step": 698
    },
    {
      "epoch": 2.7888446215139444,
      "grad_norm": 0.4687785804271698,
      "learning_rate": 1.5446402234531934e-06,
      "loss": 1.0225905179977417,
      "step": 700
    },
    {
      "epoch": 2.7968127490039842,
      "grad_norm": 0.43158504366874695,
      "learning_rate": 1.5321403798699924e-06,
      "loss": 0.6731637716293335,
      "step": 702
    },
    {
      "epoch": 2.804780876494024,
      "grad_norm": 0.9083545207977295,
      "learning_rate": 1.5196934745383456e-06,
      "loss": 1.2145440578460693,
      "step": 704
    },
    {
      "epoch": 2.812749003984064,
      "grad_norm": 0.3637361228466034,
      "learning_rate": 1.5073000485037044e-06,
      "loss": 1.2721225023269653,
      "step": 706
    },
    {
      "epoch": 2.8207171314741037,
      "grad_norm": 0.6472439765930176,
      "learning_rate": 1.4949606404868638e-06,
      "loss": 1.076061487197876,
      "step": 708
    },
    {
      "epoch": 2.8286852589641436,
      "grad_norm": 0.31598398089408875,
      "learning_rate": 1.4826757868605519e-06,
      "loss": 0.5194457173347473,
      "step": 710
    },
    {
      "epoch": 2.8366533864541834,
      "grad_norm": 0.422805517911911,
      "learning_rate": 1.4704460216261125e-06,
      "loss": 1.0148996114730835,
      "step": 712
    },
    {
      "epoch": 2.8446215139442232,
      "grad_norm": 0.35191410779953003,
      "learning_rate": 1.458271876390287e-06,
      "loss": 0.5827285051345825,
      "step": 714
    },
    {
      "epoch": 2.852589641434263,
      "grad_norm": 0.5099374651908875,
      "learning_rate": 1.446153880342117e-06,
      "loss": 0.977530837059021,
      "step": 716
    },
    {
      "epoch": 2.860557768924303,
      "grad_norm": 0.9145219922065735,
      "learning_rate": 1.4340925602299337e-06,
      "loss": 0.6681551933288574,
      "step": 718
    },
    {
      "epoch": 2.8685258964143427,
      "grad_norm": 0.41111892461776733,
      "learning_rate": 1.4220884403384656e-06,
      "loss": 0.5638666152954102,
      "step": 720
    },
    {
      "epoch": 2.8764940239043826,
      "grad_norm": 1.1720389127731323,
      "learning_rate": 1.4101420424660433e-06,
      "loss": 0.785491943359375,
      "step": 722
    },
    {
      "epoch": 2.8844621513944224,
      "grad_norm": 1.3605897426605225,
      "learning_rate": 1.3982538859019219e-06,
      "loss": 0.6331469416618347,
      "step": 724
    },
    {
      "epoch": 2.8924302788844622,
      "grad_norm": 0.5455276370048523,
      "learning_rate": 1.3864244874037096e-06,
      "loss": 0.9611477255821228,
      "step": 726
    },
    {
      "epoch": 2.900398406374502,
      "grad_norm": 0.5119454264640808,
      "learning_rate": 1.3746543611749007e-06,
      "loss": 0.5628127455711365,
      "step": 728
    },
    {
      "epoch": 2.908366533864542,
      "grad_norm": 0.6195021271705627,
      "learning_rate": 1.3629440188425285e-06,
      "loss": 0.9669933915138245,
      "step": 730
    },
    {
      "epoch": 2.9163346613545817,
      "grad_norm": 0.5482296347618103,
      "learning_rate": 1.351293969434921e-06,
      "loss": 0.4801521301269531,
      "step": 732
    },
    {
      "epoch": 2.9243027888446216,
      "grad_norm": 1.5037771463394165,
      "learning_rate": 1.3397047193595809e-06,
      "loss": 1.0168917179107666,
      "step": 734
    },
    {
      "epoch": 2.9322709163346614,
      "grad_norm": 1.3908429145812988,
      "learning_rate": 1.3281767723811651e-06,
      "loss": 0.45951351523399353,
      "step": 736
    },
    {
      "epoch": 2.9402390438247012,
      "grad_norm": 1.2788947820663452,
      "learning_rate": 1.3167106295995923e-06,
      "loss": 0.9385776519775391,
      "step": 738
    },
    {
      "epoch": 2.948207171314741,
      "grad_norm": 0.4047071933746338,
      "learning_rate": 1.3053067894282624e-06,
      "loss": 1.1306208372116089,
      "step": 740
    },
    {
      "epoch": 2.956175298804781,
      "grad_norm": 0.6707137823104858,
      "learning_rate": 1.2939657475723844e-06,
      "loss": 0.4726404547691345,
      "step": 742
    },
    {
      "epoch": 2.9641434262948207,
      "grad_norm": 0.41260018944740295,
      "learning_rate": 1.2826879970074346e-06,
      "loss": 0.9435449838638306,
      "step": 744
    },
    {
      "epoch": 2.9721115537848606,
      "grad_norm": 0.4861120581626892,
      "learning_rate": 1.271474027957727e-06,
      "loss": 0.8665035367012024,
      "step": 746
    },
    {
      "epoch": 2.9800796812749004,
      "grad_norm": 0.36213749647140503,
      "learning_rate": 1.2603243278751043e-06,
      "loss": 0.5380903482437134,
      "step": 748
    },
    {
      "epoch": 2.9880478087649402,
      "grad_norm": 0.28600847721099854,
      "learning_rate": 1.249239381417746e-06,
      "loss": 0.14900051057338715,
      "step": 750
    },
    {
      "epoch": 2.99601593625498,
      "grad_norm": 0.2858439087867737,
      "learning_rate": 1.238219670429104e-06,
      "loss": 0.9874935150146484,
      "step": 752
    },
    {
      "epoch": 3.00398406374502,
      "grad_norm": 0.3885032832622528,
      "learning_rate": 1.2272656739169589e-06,
      "loss": 0.5676672458648682,
      "step": 754
    },
    {
      "epoch": 3.0119521912350598,
      "grad_norm": 0.5316081047058105,
      "learning_rate": 1.2163778680325948e-06,
      "loss": 0.5932224988937378,
      "step": 756
    },
    {
      "epoch": 3.0199203187250996,
      "grad_norm": 1.6705152988433838,
      "learning_rate": 1.2055567260501055e-06,
      "loss": 0.38578367233276367,
      "step": 758
    },
    {
      "epoch": 3.0278884462151394,
      "grad_norm": 0.453227162361145,
      "learning_rate": 1.1948027183458182e-06,
      "loss": 0.6911048889160156,
      "step": 760
    },
    {
      "epoch": 3.0358565737051793,
      "grad_norm": 0.5900776982307434,
      "learning_rate": 1.18411631237785e-06,
      "loss": 0.49544647336006165,
      "step": 762
    },
    {
      "epoch": 3.043824701195219,
      "grad_norm": 0.41954505443573,
      "learning_rate": 1.173497972665788e-06,
      "loss": 0.9106304049491882,
      "step": 764
    },
    {
      "epoch": 3.051792828685259,
      "grad_norm": 0.6160275340080261,
      "learning_rate": 1.1629481607704964e-06,
      "loss": 0.17725174129009247,
      "step": 766
    },
    {
      "epoch": 3.0597609561752988,
      "grad_norm": 0.3419613540172577,
      "learning_rate": 1.1524673352740543e-06,
      "loss": 0.5901432037353516,
      "step": 768
    },
    {
      "epoch": 3.0677290836653386,
      "grad_norm": 0.4160565733909607,
      "learning_rate": 1.1420559517598206e-06,
      "loss": 0.7431409358978271,
      "step": 770
    },
    {
      "epoch": 3.0756972111553784,
      "grad_norm": 0.4542829394340515,
      "learning_rate": 1.13171446279263e-06,
      "loss": 0.8950256705284119,
      "step": 772
    },
    {
      "epoch": 3.0836653386454183,
      "grad_norm": 0.9215657711029053,
      "learning_rate": 1.1214433178991258e-06,
      "loss": 0.8565894365310669,
      "step": 774
    },
    {
      "epoch": 3.091633466135458,
      "grad_norm": 0.11135963350534439,
      "learning_rate": 1.111242963548212e-06,
      "loss": 0.5183956623077393,
      "step": 776
    },
    {
      "epoch": 3.099601593625498,
      "grad_norm": 0.7540588974952698,
      "learning_rate": 1.1011138431316538e-06,
      "loss": 0.6138868927955627,
      "step": 778
    },
    {
      "epoch": 3.1075697211155378,
      "grad_norm": 0.4166029095649719,
      "learning_rate": 1.0910563969447969e-06,
      "loss": 0.8939814567565918,
      "step": 780
    },
    {
      "epoch": 3.1155378486055776,
      "grad_norm": 0.16396933794021606,
      "learning_rate": 1.0810710621674355e-06,
      "loss": 0.36445876955986023,
      "step": 782
    },
    {
      "epoch": 3.1235059760956174,
      "grad_norm": 0.5422102212905884,
      "learning_rate": 1.0711582728448034e-06,
      "loss": 0.9743568897247314,
      "step": 784
    },
    {
      "epoch": 3.1314741035856573,
      "grad_norm": 0.35944244265556335,
      "learning_rate": 1.0613184598687088e-06,
      "loss": 0.7615415453910828,
      "step": 786
    },
    {
      "epoch": 3.139442231075697,
      "grad_norm": 1.8137507438659668,
      "learning_rate": 1.0515520509588076e-06,
      "loss": 0.6376831531524658,
      "step": 788
    },
    {
      "epoch": 3.147410358565737,
      "grad_norm": 0.3689141571521759,
      "learning_rate": 1.0418594706440022e-06,
      "loss": 0.4886000454425812,
      "step": 790
    },
    {
      "epoch": 3.1553784860557768,
      "grad_norm": 0.8319191932678223,
      "learning_rate": 1.0322411402439982e-06,
      "loss": 0.8063745498657227,
      "step": 792
    },
    {
      "epoch": 3.1633466135458166,
      "grad_norm": 0.14646872878074646,
      "learning_rate": 1.0226974778509846e-06,
      "loss": 0.22115838527679443,
      "step": 794
    },
    {
      "epoch": 3.1713147410358564,
      "grad_norm": 0.6124083995819092,
      "learning_rate": 1.0132288983114611e-06,
      "loss": 0.5966731905937195,
      "step": 796
    },
    {
      "epoch": 3.1792828685258963,
      "grad_norm": 1.614851951599121,
      "learning_rate": 1.003835813208206e-06,
      "loss": 0.5179063677787781,
      "step": 798
    },
    {
      "epoch": 3.187250996015936,
      "grad_norm": 8.475183486938477,
      "learning_rate": 9.945186308423832e-07,
      "loss": 0.8425936102867126,
      "step": 800
    },
    {
      "epoch": 3.195219123505976,
      "grad_norm": 0.23035867512226105,
      "learning_rate": 9.852777562157983e-07,
      "loss": 0.6974341869354248,
      "step": 802
    },
    {
      "epoch": 3.2031872509960158,
      "grad_norm": 1.2881801128387451,
      "learning_rate": 9.7611359101329e-07,
      "loss": 0.7119286060333252,
      "step": 804
    },
    {
      "epoch": 3.2111553784860556,
      "grad_norm": 0.9506440758705139,
      "learning_rate": 9.670265335852725e-07,
      "loss": 1.1356767416000366,
      "step": 806
    },
    {
      "epoch": 3.2191235059760954,
      "grad_norm": 1.7327920198440552,
      "learning_rate": 9.580169789304152e-07,
      "loss": 0.48909544944763184,
      "step": 808
    },
    {
      "epoch": 3.2270916334661353,
      "grad_norm": 0.6548786163330078,
      "learning_rate": 9.490853186784801e-07,
      "loss": 0.6980054974555969,
      "step": 810
    },
    {
      "epoch": 3.235059760956175,
      "grad_norm": 0.7706406116485596,
      "learning_rate": 9.402319410732911e-07,
      "loss": 0.37898892164230347,
      "step": 812
    },
    {
      "epoch": 3.243027888446215,
      "grad_norm": 0.4636201858520508,
      "learning_rate": 9.314572309558617e-07,
      "loss": 0.8463274836540222,
      "step": 814
    },
    {
      "epoch": 3.2509960159362548,
      "grad_norm": 1.491005778312683,
      "learning_rate": 9.227615697476674e-07,
      "loss": 0.2890554964542389,
      "step": 816
    },
    {
      "epoch": 3.2589641434262946,
      "grad_norm": 0.49871179461479187,
      "learning_rate": 9.141453354340619e-07,
      "loss": 0.8816354274749756,
      "step": 818
    },
    {
      "epoch": 3.2669322709163344,
      "grad_norm": 1.8240610361099243,
      "learning_rate": 9.056089025478495e-07,
      "loss": 0.4744340479373932,
      "step": 820
    },
    {
      "epoch": 3.2749003984063743,
      "grad_norm": 1.593883752822876,
      "learning_rate": 8.971526421530067e-07,
      "loss": 0.8082240223884583,
      "step": 822
    },
    {
      "epoch": 3.2828685258964145,
      "grad_norm": 0.8863897919654846,
      "learning_rate": 8.887769218285469e-07,
      "loss": 0.5582653284072876,
      "step": 824
    },
    {
      "epoch": 3.2908366533864544,
      "grad_norm": 0.3828669786453247,
      "learning_rate": 8.804821056525498e-07,
      "loss": 0.9310887455940247,
      "step": 826
    },
    {
      "epoch": 3.298804780876494,
      "grad_norm": 0.7599470019340515,
      "learning_rate": 8.722685541863284e-07,
      "loss": 0.9080052375793457,
      "step": 828
    },
    {
      "epoch": 3.306772908366534,
      "grad_norm": 0.6911408305168152,
      "learning_rate": 8.641366244587612e-07,
      "loss": 0.843339204788208,
      "step": 830
    },
    {
      "epoch": 3.314741035856574,
      "grad_norm": 1.4359130859375,
      "learning_rate": 8.560866699507703e-07,
      "loss": 0.5319377183914185,
      "step": 832
    },
    {
      "epoch": 3.3227091633466137,
      "grad_norm": 0.7760400176048279,
      "learning_rate": 8.481190405799569e-07,
      "loss": 0.3904685080051422,
      "step": 834
    },
    {
      "epoch": 3.3306772908366535,
      "grad_norm": 1.4175999164581299,
      "learning_rate": 8.402340826853908e-07,
      "loss": 0.557598352432251,
      "step": 836
    },
    {
      "epoch": 3.3386454183266934,
      "grad_norm": 0.6159012317657471,
      "learning_rate": 8.324321390125548e-07,
      "loss": 0.7651152014732361,
      "step": 838
    },
    {
      "epoch": 3.346613545816733,
      "grad_norm": 1.8655797243118286,
      "learning_rate": 8.247135486984487e-07,
      "loss": 0.25711414217948914,
      "step": 840
    },
    {
      "epoch": 3.354581673306773,
      "grad_norm": 1.0708720684051514,
      "learning_rate": 8.170786472568451e-07,
      "loss": 0.7081012725830078,
      "step": 842
    },
    {
      "epoch": 3.362549800796813,
      "grad_norm": 0.5424862504005432,
      "learning_rate": 8.095277665637061e-07,
      "loss": 0.675047755241394,
      "step": 844
    },
    {
      "epoch": 3.3705179282868527,
      "grad_norm": 0.34637734293937683,
      "learning_rate": 8.020612348427572e-07,
      "loss": 0.9298898577690125,
      "step": 846
    },
    {
      "epoch": 3.3784860557768925,
      "grad_norm": 0.4564300775527954,
      "learning_rate": 7.946793766512195e-07,
      "loss": 0.8776551485061646,
      "step": 848
    },
    {
      "epoch": 3.3864541832669324,
      "grad_norm": 1.1965371370315552,
      "learning_rate": 7.873825128657034e-07,
      "loss": 1.0951646566390991,
      "step": 850
    },
    {
      "epoch": 3.394422310756972,
      "grad_norm": 0.3990395665168762,
      "learning_rate": 7.801709606682573e-07,
      "loss": 0.8855547308921814,
      "step": 852
    },
    {
      "epoch": 3.402390438247012,
      "grad_norm": 0.5456656217575073,
      "learning_rate": 7.730450335325852e-07,
      "loss": 0.5132330060005188,
      "step": 854
    },
    {
      "epoch": 3.410358565737052,
      "grad_norm": 0.9289339780807495,
      "learning_rate": 7.660050412104153e-07,
      "loss": 0.8418811559677124,
      "step": 856
    },
    {
      "epoch": 3.4183266932270917,
      "grad_norm": 0.7988582849502563,
      "learning_rate": 7.5905128971804e-07,
      "loss": 1.006464958190918,
      "step": 858
    },
    {
      "epoch": 3.4262948207171315,
      "grad_norm": 0.6170944571495056,
      "learning_rate": 7.521840813230098e-07,
      "loss": 1.038280963897705,
      "step": 860
    },
    {
      "epoch": 3.4342629482071714,
      "grad_norm": 0.7820700407028198,
      "learning_rate": 7.454037145309986e-07,
      "loss": 0.660965085029602,
      "step": 862
    },
    {
      "epoch": 3.442231075697211,
      "grad_norm": 1.3111145496368408,
      "learning_rate": 7.38710484072826e-07,
      "loss": 0.5856328010559082,
      "step": 864
    },
    {
      "epoch": 3.450199203187251,
      "grad_norm": 0.4656525254249573,
      "learning_rate": 7.321046808916427e-07,
      "loss": 0.8252797722816467,
      "step": 866
    },
    {
      "epoch": 3.458167330677291,
      "grad_norm": 0.43764209747314453,
      "learning_rate": 7.255865921302902e-07,
      "loss": 0.7282843589782715,
      "step": 868
    },
    {
      "epoch": 3.4661354581673307,
      "grad_norm": 0.6524649858474731,
      "learning_rate": 7.191565011188149e-07,
      "loss": 0.5070784091949463,
      "step": 870
    },
    {
      "epoch": 3.4741035856573705,
      "grad_norm": 0.3002372086048126,
      "learning_rate": 7.128146873621533e-07,
      "loss": 0.5543745756149292,
      "step": 872
    },
    {
      "epoch": 3.4820717131474104,
      "grad_norm": 0.2070620357990265,
      "learning_rate": 7.065614265279819e-07,
      "loss": 0.613795816898346,
      "step": 874
    },
    {
      "epoch": 3.49003984063745,
      "grad_norm": 0.42765840888023376,
      "learning_rate": 7.00396990434734e-07,
      "loss": 0.5247716903686523,
      "step": 876
    },
    {
      "epoch": 3.49800796812749,
      "grad_norm": 0.6999766230583191,
      "learning_rate": 6.943216470397872e-07,
      "loss": 0.5478222370147705,
      "step": 878
    },
    {
      "epoch": 3.50597609561753,
      "grad_norm": 0.20102103054523468,
      "learning_rate": 6.883356604278116e-07,
      "loss": 0.3919145166873932,
      "step": 880
    },
    {
      "epoch": 3.5139442231075697,
      "grad_norm": 0.4787895977497101,
      "learning_rate": 6.824392907992938e-07,
      "loss": 0.5872817635536194,
      "step": 882
    },
    {
      "epoch": 3.5219123505976095,
      "grad_norm": 1.4848010540008545,
      "learning_rate": 6.76632794459224e-07,
      "loss": 0.19930166006088257,
      "step": 884
    },
    {
      "epoch": 3.5298804780876494,
      "grad_norm": 0.8242633938789368,
      "learning_rate": 6.709164238059568e-07,
      "loss": 0.7871568202972412,
      "step": 886
    },
    {
      "epoch": 3.537848605577689,
      "grad_norm": 0.6071775555610657,
      "learning_rate": 6.652904273202385e-07,
      "loss": 0.8753920793533325,
      "step": 888
    },
    {
      "epoch": 3.545816733067729,
      "grad_norm": 0.45464256405830383,
      "learning_rate": 6.597550495544071e-07,
      "loss": 0.4655469059944153,
      "step": 890
    },
    {
      "epoch": 3.553784860557769,
      "grad_norm": 0.8208728432655334,
      "learning_rate": 6.54310531121762e-07,
      "loss": 1.0298000574111938,
      "step": 892
    },
    {
      "epoch": 3.5617529880478087,
      "grad_norm": 0.48669975996017456,
      "learning_rate": 6.489571086861027e-07,
      "loss": 1.0191948413848877,
      "step": 894
    },
    {
      "epoch": 3.5697211155378485,
      "grad_norm": 0.31976720690727234,
      "learning_rate": 6.436950149514448e-07,
      "loss": 0.21331249177455902,
      "step": 896
    },
    {
      "epoch": 3.5776892430278884,
      "grad_norm": 0.37387582659721375,
      "learning_rate": 6.385244786519036e-07,
      "loss": 0.6412251591682434,
      "step": 898
    },
    {
      "epoch": 3.585657370517928,
      "grad_norm": 0.590537428855896,
      "learning_rate": 6.334457245417491e-07,
      "loss": 0.2700989246368408,
      "step": 900
    },
    {
      "epoch": 3.593625498007968,
      "grad_norm": 0.6956413984298706,
      "learning_rate": 6.284589733856403e-07,
      "loss": 0.10628986358642578,
      "step": 902
    },
    {
      "epoch": 3.601593625498008,
      "grad_norm": 0.49798694252967834,
      "learning_rate": 6.235644419490253e-07,
      "loss": 0.7536365389823914,
      "step": 904
    },
    {
      "epoch": 3.6095617529880477,
      "grad_norm": 0.780397891998291,
      "learning_rate": 6.18762342988722e-07,
      "loss": 0.6219207644462585,
      "step": 906
    },
    {
      "epoch": 3.6175298804780875,
      "grad_norm": 0.42463183403015137,
      "learning_rate": 6.140528852436666e-07,
      "loss": 0.8660494089126587,
      "step": 908
    },
    {
      "epoch": 3.6254980079681274,
      "grad_norm": 0.8319142460823059,
      "learning_rate": 6.094362734258438e-07,
      "loss": 0.8276482224464417,
      "step": 910
    },
    {
      "epoch": 3.633466135458167,
      "grad_norm": 0.3002433776855469,
      "learning_rate": 6.049127082113853e-07,
      "loss": 0.04957403242588043,
      "step": 912
    },
    {
      "epoch": 3.641434262948207,
      "grad_norm": 0.6337653994560242,
      "learning_rate": 6.004823862318475e-07,
      "loss": 0.7774660587310791,
      "step": 914
    },
    {
      "epoch": 3.649402390438247,
      "grad_norm": 0.9938945770263672,
      "learning_rate": 5.961455000656656e-07,
      "loss": 0.8184969425201416,
      "step": 916
    },
    {
      "epoch": 3.6573705179282867,
      "grad_norm": 2.279283285140991,
      "learning_rate": 5.919022382297808e-07,
      "loss": 0.18306171894073486,
      "step": 918
    },
    {
      "epoch": 3.6653386454183265,
      "grad_norm": 0.7640562653541565,
      "learning_rate": 5.877527851714479e-07,
      "loss": 0.9448274970054626,
      "step": 920
    },
    {
      "epoch": 3.6733067729083664,
      "grad_norm": 0.46138614416122437,
      "learning_rate": 5.836973212602144e-07,
      "loss": 0.13672778010368347,
      "step": 922
    },
    {
      "epoch": 3.681274900398406,
      "grad_norm": 1.4605869054794312,
      "learning_rate": 5.797360227800832e-07,
      "loss": 1.0739829540252686,
      "step": 924
    },
    {
      "epoch": 3.6892430278884465,
      "grad_norm": 1.0049115419387817,
      "learning_rate": 5.758690619218493e-07,
      "loss": 0.32028231024742126,
      "step": 926
    },
    {
      "epoch": 3.6972111553784863,
      "grad_norm": 0.27289578318595886,
      "learning_rate": 5.720966067756131e-07,
      "loss": 0.5045006275177002,
      "step": 928
    },
    {
      "epoch": 3.705179282868526,
      "grad_norm": 0.21244530379772186,
      "learning_rate": 5.684188213234763e-07,
      "loss": 0.775736927986145,
      "step": 930
    },
    {
      "epoch": 3.713147410358566,
      "grad_norm": 0.4776446223258972,
      "learning_rate": 5.648358654324119e-07,
      "loss": 0.89571613073349,
      "step": 932
    },
    {
      "epoch": 3.721115537848606,
      "grad_norm": 2.134922981262207,
      "learning_rate": 5.613478948473166e-07,
      "loss": 0.5767644047737122,
      "step": 934
    },
    {
      "epoch": 3.7290836653386457,
      "grad_norm": 0.4431370794773102,
      "learning_rate": 5.579550611842395e-07,
      "loss": 0.8775071501731873,
      "step": 936
    },
    {
      "epoch": 3.7370517928286855,
      "grad_norm": 1.2038134336471558,
      "learning_rate": 5.546575119237923e-07,
      "loss": 0.7694797515869141,
      "step": 938
    },
    {
      "epoch": 3.7450199203187253,
      "grad_norm": 0.5099160671234131,
      "learning_rate": 5.514553904047394e-07,
      "loss": 0.7644805312156677,
      "step": 940
    },
    {
      "epoch": 3.752988047808765,
      "grad_norm": 1.2608742713928223,
      "learning_rate": 5.483488358177647e-07,
      "loss": 0.8039394617080688,
      "step": 942
    },
    {
      "epoch": 3.760956175298805,
      "grad_norm": 0.5588910579681396,
      "learning_rate": 5.453379831994238e-07,
      "loss": 0.5133582949638367,
      "step": 944
    },
    {
      "epoch": 3.768924302788845,
      "grad_norm": 0.4913143813610077,
      "learning_rate": 5.424229634262735e-07,
      "loss": 0.9352614283561707,
      "step": 946
    },
    {
      "epoch": 3.7768924302788847,
      "grad_norm": 0.9081344604492188,
      "learning_rate": 5.396039032091826e-07,
      "loss": 0.5901059508323669,
      "step": 948
    },
    {
      "epoch": 3.7848605577689245,
      "grad_norm": 0.3207649290561676,
      "learning_rate": 5.368809250878228e-07,
      "loss": 0.9115123748779297,
      "step": 950
    },
    {
      "epoch": 3.7928286852589643,
      "grad_norm": 0.37404128909111023,
      "learning_rate": 5.342541474253438e-07,
      "loss": 0.8896762132644653,
      "step": 952
    },
    {
      "epoch": 3.800796812749004,
      "grad_norm": 0.8347617387771606,
      "learning_rate": 5.317236844032285e-07,
      "loss": 0.5753062963485718,
      "step": 954
    },
    {
      "epoch": 3.808764940239044,
      "grad_norm": 0.7320111989974976,
      "learning_rate": 5.292896460163282e-07,
      "loss": 0.7715502977371216,
      "step": 956
    },
    {
      "epoch": 3.816733067729084,
      "grad_norm": 2.449096918106079,
      "learning_rate": 5.269521380680823e-07,
      "loss": 0.5903229713439941,
      "step": 958
    },
    {
      "epoch": 3.8247011952191237,
      "grad_norm": 0.4072583317756653,
      "learning_rate": 5.247112621659186e-07,
      "loss": 0.761043906211853,
      "step": 960
    },
    {
      "epoch": 3.8326693227091635,
      "grad_norm": 0.32995879650115967,
      "learning_rate": 5.225671157168378e-07,
      "loss": 0.5542949438095093,
      "step": 962
    },
    {
      "epoch": 3.8406374501992033,
      "grad_norm": 0.5266864895820618,
      "learning_rate": 5.205197919231773e-07,
      "loss": 0.93712317943573,
      "step": 964
    },
    {
      "epoch": 3.848605577689243,
      "grad_norm": 0.5685020089149475,
      "learning_rate": 5.185693797785624e-07,
      "loss": 0.35376280546188354,
      "step": 966
    },
    {
      "epoch": 3.856573705179283,
      "grad_norm": 0.6125916838645935,
      "learning_rate": 5.167159640640358e-07,
      "loss": 0.5749203562736511,
      "step": 968
    },
    {
      "epoch": 3.864541832669323,
      "grad_norm": 1.0156006813049316,
      "learning_rate": 5.149596253443733e-07,
      "loss": 0.6873950958251953,
      "step": 970
    },
    {
      "epoch": 3.8725099601593627,
      "grad_norm": 0.17581939697265625,
      "learning_rate": 5.133004399645818e-07,
      "loss": 0.5502765774726868,
      "step": 972
    },
    {
      "epoch": 3.8804780876494025,
      "grad_norm": 0.16612528264522552,
      "learning_rate": 5.117384800465806e-07,
      "loss": 0.44509047269821167,
      "step": 974
    },
    {
      "epoch": 3.8884462151394423,
      "grad_norm": 0.38767173886299133,
      "learning_rate": 5.102738134860656e-07,
      "loss": 0.8849189877510071,
      "step": 976
    },
    {
      "epoch": 3.896414342629482,
      "grad_norm": 0.660287082195282,
      "learning_rate": 5.089065039495598e-07,
      "loss": 0.5235478281974792,
      "step": 978
    },
    {
      "epoch": 3.904382470119522,
      "grad_norm": 0.180130273103714,
      "learning_rate": 5.076366108716438e-07,
      "loss": 0.41882917284965515,
      "step": 980
    },
    {
      "epoch": 3.912350597609562,
      "grad_norm": 0.6282052993774414,
      "learning_rate": 5.064641894523737e-07,
      "loss": 0.45488396286964417,
      "step": 982
    },
    {
      "epoch": 3.9203187250996017,
      "grad_norm": 0.450431227684021,
      "learning_rate": 5.053892906548811e-07,
      "loss": 0.8408481478691101,
      "step": 984
    },
    {
      "epoch": 3.9282868525896415,
      "grad_norm": 0.8664645552635193,
      "learning_rate": 5.044119612031576e-07,
      "loss": 0.18552815914154053,
      "step": 986
    },
    {
      "epoch": 3.9362549800796813,
      "grad_norm": 1.059046745300293,
      "learning_rate": 5.035322435800248e-07,
      "loss": 0.5966424942016602,
      "step": 988
    },
    {
      "epoch": 3.944223107569721,
      "grad_norm": 0.42316704988479614,
      "learning_rate": 5.027501760252857e-07,
      "loss": 0.5332047343254089,
      "step": 990
    },
    {
      "epoch": 3.952191235059761,
      "grad_norm": 0.8692534565925598,
      "learning_rate": 5.02065792534065e-07,
      "loss": 0.9967976212501526,
      "step": 992
    },
    {
      "epoch": 3.960159362549801,
      "grad_norm": 0.054123349487781525,
      "learning_rate": 5.014791228553294e-07,
      "loss": 0.29197585582733154,
      "step": 994
    },
    {
      "epoch": 3.9681274900398407,
      "grad_norm": 0.6394121050834656,
      "learning_rate": 5.00990192490595e-07,
      "loss": 0.8136619925498962,
      "step": 996
    },
    {
      "epoch": 3.9760956175298805,
      "grad_norm": 0.3882506191730499,
      "learning_rate": 5.005990226928196e-07,
      "loss": 0.5231336951255798,
      "step": 998
    },
    {
      "epoch": 3.9840637450199203,
      "grad_norm": 0.547065019607544,
      "learning_rate": 5.003056304654774e-07,
      "loss": 0.4203438460826874,
      "step": 1000
    },
    {
      "epoch": 3.99203187250996,
      "grad_norm": 0.6402074098587036,
      "learning_rate": 5.001100285618214e-07,
      "loss": 0.8936667442321777,
      "step": 1002
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.8004528284072876,
      "learning_rate": 5.000122254843282e-07,
      "loss": 0.2594180703163147,
      "step": 1004
    },
    {
      "epoch": 4.0,
      "step": 1004,
      "total_flos": 4.038502240003031e+18,
      "train_loss": 0.8685887347743568,
      "train_runtime": 10574.1516,
      "train_samples_per_second": 5.697,
      "train_steps_per_second": 0.095
    }
  ],
  "logging_steps": 2,
  "max_steps": 1004,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 99999,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.038502240003031e+18,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}